Pseudoalteromonas spp. Serve as Initial Bacterial Attractants in Mesocosms of Coastal Waters but Have Subsequent Antifouling Capacity in Mesocosms and when Embedded in Paint

Abstract

The purpose of the present study was to determine if the monoculture antifouling effect of several pigmented pseudoalteromonads was retained in in vitro mesocosm systems using natural coastal seawater and when the bacteria were embedded in paint used on surfaces submerged in coastal waters. Pseudoalteromonas piscicida survived on a steel surface and retained antifouling activity for at least 53 days in sterile seawater, whereas P. tunicata survived and had antifouling activity for only 1 week. However, during the first week, all Pseudoalteromonas strains facilitated rather than prevented bacterial attachment when used to coat stainless steel surfaces and submerged in mesocosms with natural seawater. The bacterial density on surfaces coated with sterile growth medium was 10⁵ cells/cm² after 7 days, whereas counts on surfaces precoated with Pseudoalteromonas were significantly higher, at 10⁶ to 10⁸ cells/cm². However, after 53 days, seven of eight Pseudoalteromonas strains had reduced total bacterial adhesion compared to the control. P. piscicida, P. antarctica, and P. ulvae remained on the surface, at levels similar to those in the initial coating, whereas P. tunicata could not be detected. Larger fouling organisms were observed on all plates precoated with Pseudoalteromonas; however, plates coated only with sterile growth medium were dominated by a bacterial biofilm. Suspensions of a P. piscicida strain and a P. tunicata strain were incorporated into ship paints (Hempasil x3 87500 and Hempasil 77500) used on plates that were placed at the Hempel A/S test site in Jyllinge Harbor. For the first 4 months, no differences were observed between control plates and treated plates, but after 5 to 6 months, the control plates were more fouled than the plates with pseudoalteromonad-based paint. Our study demonstrates that no single laboratory assay can predict antifouling effects and that a combination of laboratory and real-life methods must be used to determine the potential antifouling capability of new agents or organisms.

INTRODUCTION

In the marine environment, artificial surfaces never remain pristine but quickly become colonized by a layer of marine bacteria (1), which serves as a settlement cue for subsequent attachment of other organisms (2, 3). Biofouling of ship hulls results in significantly increased fuel consumption and corrosion, causing increased costs and CO₂ emissions (4). Biofouling can also serve as a carrier and introduce invasive marine organisms into new environments (4), and costly mechanical processes coupled with toxic heavy metal-based paint have been used to combat marine biofouling (4). An example is tributyltin, which is believed to contribute to development of antimicrobial tolerance in marine organisms (5) and to cause imposex in some invertebrates (6–8). The International Maritime Organization also reported its accumulation in mammals and debilitation of immunological defenses in fish (9). As a result, antifouling paints containing tin were banned in 2003 and gradually removed from shipping fleets (10). Consequently, the search for environmentally friendly antifouling compounds or principles that reduce or eliminate attachment of marine organisms is intense (11).

A number of studies have suggested that naturally occurring marine macro- and microorganisms and their production of secondary metabolites could be used as an antifouling strategy with limited impact on the marine ecology (11–13). Marine macroorganisms produce antifouling compounds such as the halogenated furanones produced by the red algae Delisea pulchra (14) and terpenoids isolated from the marine sponge Acanthella cavernosa (15), but they also rely on epiphytic bacteria as producers of antifouling compounds (16–18). From a biotechnological perspective, microorganisms are an exploitable source of antifouling compounds (19), and we and other researchers have previously shown that production of bioactive, antagonistic compounds is common among marine bacteria (20, 21).

Several studies on antifouling agents from marine bacteria have focused on the genus Pseudoalteromonas, since several species of this genus produce bioactive compounds that target a broad range of marine biofilms (22) and fouling organisms (23, 24). Some Pseudoalteromonas compounds have antibacterial, algicidal, antifungal, or antiviral activity (24). A well-studied member of the genus Pseudoalteromonas is the green-pigmented bacterium Pseudoalteromonas tunicata (23), which produces at least five different extracellular compounds that specifically inhibit the growth and settlement of different classes of marine organisms (25). Pseudoalteromonads are very successful colonizers of biotic surfaces and are often associated with eukaryotes (22), and it is hypothesized that their wide spectrum of bioactive compounds serves in part to protect their host from fouling (26). One key component of the P. tunicata bioactive compounds is the AlpP protein, which has both antibacterial and autolytic activities (27) due to its ability to generate hydrogen peroxide (28). A quantitative PCR (qPCR) based on the alpP gene has been developed for specific quantification of P. tunicata (29).

We recently demonstrated that not only P. tunicata but also other species within the Pseudoalteromonas genus prevented adhesion of the marine fouling bacterium Pseudoalteromonas S91 and spores of the green alga Ulva australis (30). Since the antifouling effect was seen against both prokaryotes and eukaryotes, we hypothesize that these strains would also repel common microbial colonizers in natural, more complex systems. The purpose of the present study was to probe the antifouling potential of P. piscicida, P. tunicata, and P. ulvae in a system mimicking the natural marine environment and also to determine the potential antifouling effect when bacteria were incorporated into ship paints and tested under real-life conditions.

MATERIALS AND METHODS

Bacterial strains and growth conditions.

Pseudoalteromonas piscicida strains B39bio and A38q-4a (yellow), P. tunicata strains J49q-3a (purple), J36q-4a (purple), and J38a-5a (yellow), P. ulvae strain H34q-5a (green), and P. antarctica strain E45q-4a (brown) were isolated in 2009 from Danish coastal waters or biotic surfaces (30). Samples were plated on marine agar and replica plated onto Vibrio-embedded agar (30), and colonies causing clearing zones in the turbid Vibrio layer were isolated, pure cultured, identified, and stored at −80°C. For the present study, cultures were revived from frozen stocks for each experiment. P. tunicata strain D2 (23) was included as a presumed positive antifouling control and was a kind gift from the University of New South Wales. Pseudoalteromonas S91 is a pronounced biofilm former (31). It is light pink and streptomycin resistant and was kindly donated by the University of Adelaide. The P. piscicida and P. tunicata strains have antibacterial and/or antifouling activity in in vitro assays (30). The bacteria were cultured at 25°C on marine agar (MA; Difco) or on half-strength MA (50% MA) and diluted in 3% Sigma sea salt (SSS; Sigma-Aldrich) or marine broth (MB; Difco).

Survival and antifouling effects of Pseudoalteromonas biofilms on stainless steel in sterile seawater.

We first determined if the Pseudoalteromonas strains would survive and maintain antifouling activity (six of the eight strains) when exposed to sterile natural seawater. Stainless steel coupons (seafood-grade AISI 316, unpolished, type 4 finish; 0.5 × 0.5 cm, with a thickness of 1 mm) were cleaned, degreased, and sterilized as previously described (32) and then placed individually in a tilted state in wells of a 96-well microtiter plate (Nunc). Pseudoalteromonas strains were cultured in marine broth for 3 days at 25°C and then diluted 1,000-fold in MB, and 200 μl was transferred to each well containing a coupon. The bacteria were left to grow and attach to the stainless steel surface for 3 days at room temperature (20°C), and the coupons were then removed by being held at the edge by tweezers and were rinsed with 2 ml sterile 3% SSS to remove nonadhering bacteria. The plates with adhered bacteria were transferred to a new 96-well microtiter plate and exposed to 200 μl sterile-filtered seawater (0.02 μm; Whatman, Maidstone, England). The water was collected at the Danish harbor Jyllinge Harbor, and coupons were submerged in seawater with and without the addition of Pseudoalteromonas strain S91. Strain S91 was pregrown in MB for 1 day, reaching 6 × 10⁸ CFU/ml, and then diluted 1,000-fold in seawater, to a concentration of 6 × 10⁵ CFU/ml. The exact cell numbers were determined by plate counts on MA plates. The steel plates were incubated under shaking conditions (50 rpm) at room temperature, and steel coupons were removed immediately (15 to 30 min) after setup of the experiment and after 7 and 53 days. Counts of surface-attached bacteria and bacterial counts in suspension were determined as described below. All adhesion assays were carried out in triplicate.

Fouling batch mesocosm system.

Biofouling development from natural coastal seawater was observed in a batch mesocosm with seawater collected at Jyllinge Harbor. Surfaces in this area foul rapidly, and the site has been chosen by the Danish paint manufacturer Hempel A/S for testing of antifouling principles. Water temperature and salinity were measured in situ with a handheld Professional Plus instrument (model YS6050000; YSI, Yellow Springs, OH). The bacterial content of the water was quantified using plate counts on marine agar and SYBR gold staining of bacteria from 5 ml seawater filtered onto 0.02-μm Anodisc filters (Whatman) (20). Cells were counted in five fields per sample, using an Olympus BX51 microscope with 460- to 490-nm excitation and >510-nm emission filters. Stainless steel coupons (seafood-grade AISI 316, unpolished, type 4 finish; 1.0 × 2.0 cm, with a thickness of 1 mm) were handled as described above and clamped vertically in stainless steel circular racks placed in beakers that were covered with foil and sterilized by autoclaving. Each rack holds up to 20 coupons in a circle and allows circulation of liquid. Two hundred fifty milliliters of seawater was added to each of two beakers with stainless steel coupons. Fouling took place on both sides of the coupons at room temperature under slow-stirring (250 rpm on a shaking table) conditions. Samples were taken after 4 h and 1, 3, 9, 20, and 30 days, and fouling on the coupons was visualized by SYBR Gold staining followed by fluorescence microscopy (BX51; Olympus) as described above.

Influence of Pseudoalteromonas strains on fouling of stainless steel submerged in seawater.

Potential antifouling Pseudoalteromonas strains (Table 1) were cultured in marine broth for 3 days at 25°C, and 1-ml samples were used to inoculate 250 ml MB in beakers containing stainless steel coupons inserted in a circular rack. Each Pseudoalteromonas strain was inoculated in a separate beaker to avoid cross-contamination. Biofilm formation took place on both sides of the coupons at room temperature under slow-stirring (250 rpm) conditions for 3 days, and the rack containing the coupons was transferred to a new sterile beaker. Two hundred fifty milliliters of seawater collected at Jyllinge Harbor (in August) was added to the beakers containing the coupons with pregrown Pseudoalteromonas biofilms. Water temperature and salinity were measured as described above. Fouling from the seawater took place on both sides of the coupons at room temperature under slow-stirring (250 rpm) conditions, and samples were taken after 7 and 53 days. We determined (i) the number of attached culturable bacteria, (ii) the total number of attached bacteria, (iii) the number of bacteria within the Pseudoalteromonas genus, and (iv) the level of (macro)fouling. The number of culturable bacteria and the total number of bacteria in the marine water surrounding the coupons were also quantified.

Table 1.

Survival and antifouling activity of Pseudoalteromonas species on stainless steel in sterile seawater from Jyllinge Harbor at room temperature (25°C)^a

Coating species or control	Strain	No. of surface-attached bacteria (log CFU/cm2)
		Coated Pseudoalteromonas on surfaces submerged in sterile seawater			Coated Pseudoalteromonas on surfaces submerged in seawater with S91		Pseudoalteromonas S91 on surfaces
		0 days	7 days	53 days	7 days	53 days	7 days	53 days
P. piscicida	B39bio	6.9 ± 0.1	4.8 ± 1.0	4.4 ± 0.2	6.3 ± 0.2	4.7 ± 0.2	2.4 ± 0.2*	<1.3***
	A38q-4a	7.2 ± 0.1	5.2 ± 0.0	5.1 ± 0.0	6.5 ± 0.1	4.6 ± 0.3	3.0 ± 0.3*	3.5 ± 0.5
P. tunicata	J49q-3a	7.5 ± 0.0	5.5 ± 0.1	3.7 ± 0.2	5.8 ± 0.0	<1.3	3.3 ± 1.3	4.5 ± 0.7
	D2	7.6 ± 0.1	5.5 ± 0.6	<1.3	5.5 ± 0.2	<1.3	4.3 ± 1.4	4.3 ± 0.6
	J36q-4a	7.6 ± 0.6	5.2 ± 0.1	<1.3	5.7 ± 0.3	<1.3	2.4 ± 0.5*	4.3 ± 0.6
	J38a-5a	7.7 ± 0.1	5.1 ± 0.0	3.9 ± 0.1	5.3 ± 0.3	<1.3	4.3 ± 0.1	4.3 ± 0.6
P. ulvae	H34q-5a	6.8 ± 0.1	4.9 ± 0.5	<1.3	6.0 ± 0.1	4.3 ± 0.5	5.7 ± 0.3	4.4 ± 0.7
P. antarctica	E45q-4a	6.8 ± 0.2	5.2 ± 0.2	<1.3	5.9 ± 0.1	<1.3	5.8 ± 0.1	4.1 ± 0.6
Controls
Noncoated surface							4.9 ± 0.2	4.2 ± 0.6
MB-coated surface							5.0 ± 0.9	4.2 ± 0.6

^aSamples were tested immediately after submerging of coated plates (time zero) and after 0, 7, and 53 days. Pseudoalteromonas S91 was enumerated on streptomycin plates. Counts of other Pseudoalteromonas species were arrived at by subtracting the S91 count from the total count on MA plates. Counts of Pseudoalteromonas strain S91 on Pseudoalteromonas-precoated surfaces were compared to those on MB-coated surfaces by using the t test. Data are means ± standard deviations. *, P < 0.05; ***, P < 0.001.

Enumeration of culturable surface-attached bacteria.

The stainless steel coupons (with attached bacteria) were immersed in polystyrene tubes (Bibby Sterilin Ltd., Stones, United Kingdom) containing 1 ml (small coupons) or 2 ml (large coupons) sterile 3% SSS solution. Nonadherent or poorly attached bacteria were removed by carefully placing the coupons (both sides) on sterile absorbent paper before removing the bacteria from the surface by ultrasonication (30) and vortexing at maximum speed for 15 s to further facilitate removal. Tenfold serial dilutions were made in sterile 3% SSS, and colony counts of the total number of adhered bacteria were determined by plating on 50% marine agar. Where relevant, plates were inspected for pigmented colonies of the eight Pseudoalteromonas strains tested. Pseudoalteromonas S91 was enumerated on 50% MA containing 400 μg/ml streptomycin. The efficiency of the detachment procedure was verified by SYBR Gold staining followed by fluorescence microscopy (30).

Enumeration of total bacterial count by a real-time PCR procedure.

The numbers of total bacterial cells adhering to the steel surfaces were determined using a quantitative PCR method with universal primers (33). Standard curves relating the threshold cycle (C_T) to CFU/ml were made using three marine bacteria, and the method was verified by comparing bacterial densities in seawater as estimated by SYBR Gold staining and by the PCR method. Genomic DNA was extracted using phenol-chloroform extraction (34). One milliliter of sample (seawater or bacterial suspension from coupons) was centrifuged at 15,000 × g for 3 min, and the pellet was subjected to bead beating in 500 μl lysis buffer (40 mM Na₂EDTA, 0.75 M sucrose, 50 mM Tris-HCl) followed by treatment with lysozyme (1 mg/ml; 30 min at 37°C), SDS (1%), and proteinase K (2 mg/ml; overnight at 56°C). DNA was purified using phenol-chloroform-isoamyl alcohol (25:24:1) and chloroform-isoamyl alcohol (24:1) extraction and precipitated with 0.6 volume of isopropanol and 0.1 volume of 3 M sodium acetate for 1 h at 20°C. DNA was pelleted by centrifugation (20 min, 20,000 × g, 4°C), washed with 70% ethanol, and resuspended in sterile MilliQ water. The DNA contents were measured on a NanoDrop spectrophotometer (model ND-1000; Saveen Werner), and samples were stored at −20°C. We used universal bacterial primers 338f (ACTCCTACGGGAGGCAGCAG) and 518R (ATTACCGCGGCTGCTGG) for a qPCR-based quantification of the total bacterial population (33). The PCR amplifications were performed from 2 μl of purified genomic DNA by using Brilliant SYBR green QPCR master mix (Stratagene, La Jolla, CA), with ROX (6-carboxy-X-rhodamine) as a reference dye, in a 15-μl reaction volume containing 0.7 μM (each) primers, with an MX3000P instrument (Stratagene, La Jolla, CA). Standard curves relating C_T to CFU/ml were based on Phaeobacter strain 27-4 (35), Pseudoalteromonas tunicata D2 (23), and Vibrio anguillarum 90-11-287 (36). The three bacterial strains were grown in MB for 3 days under aerobic conditions and then serially diluted. DNAs were extracted from duplicate samples for each of the 10-fold dilutions, and C_T values were determined by the real-time PCR procedure. C_T values and log CFU/ml (determined on marine agar) for each dilution step were compared by linear regression. The melting curve analyses for all amplified DNA fragments were performed at temperatures from 35 to 95°C, and all showed a single peak.

Enumeration of Pseudoalteromonas and of bacteria containing the alpP gene.

All PCR amplifications were performed as described above, using 2 μl of diluted purified genomic DNA and Brilliant SYBR green QPCR master mix (Stratagene, La Jolla, CA), with ROX as a reference dye, in a 15-μl reaction volume containing 0.7 μM (each) primers, with an MX3000P instrument (Stratagene, La Jolla, CA). The number of Pseudoalteromonas organisms and the number of bacteria harboring the alpP gene were quantified using real-time PCR with previously described primers (25, 29). The PCR amplification conditions for quantification of Pseudoalteromonas were as follows: 10 min at 95°C and then 40 cycles of 30 s at 95°C, 60 s at 56°C, and 60 s at 72°C. The PCR amplification conditions for the alpP gene were as follows: 15 min at 95°C and then 40 cycles of 30 s at 92°C, 60 s at 55°C, and 60 s at 72°C, followed by a final extension of 5 min at 72°C. The melting curve analyses for all amplified DNA fragments were performed at temperatures from 35 to 95°C. Linear correlations for the number of Pseudoalteromonas organisms and the number of bacteria harboring the alpP gene were calculated based on 10-fold dilutions of P. tunicata D2 and real-time PCRs with Pseudoalteromonas (25)- and alpP (29)-specific primers, respectively. The linear regressions and correlation coefficients were as follows: log CFU/sample = −3.38 × C_T + 40.79 (R² = 0.99) and log CFU/sample = −3.55 × C_T + 46.14 (R² = 0.98).

Estimation of degree of fouling in mesocosms.

The amount of fouling on stainless steel was detected by fluorescence microscopy (magnification, ×10,000) after staining of the coupons with 0.25% SYBR Gold (Invitrogen) for 15 min in darkness. The surface was examined by direct fluorescence microscopy (Olympus BX51 fluorescence microscope) with a 460- to 490-nm excitation filter and a >510-nm barrier filter.

Antibacterial and antifouling activities of Pseudoalteromonas in ship paint.

Two-hundred-microliter aliquots of outgrown cultures of Pseudoalteromonas piscicida B39bio and P. tunicata J49q-3a were plated on marine agar and incubated for 3 days at 25°C to achieve dense growth. One milliliter of 1% Sigma sea salt was added to each MA plate, and the bacteria were removed in a thick suspension by use of a sterile Drigalski spatula. The number of bacteria in suspension was determined by dilution in 3% SSS followed by plating on marine agar, and it reached 2 × 10⁹ CFU/ml. Seventy-five microliters of bacterial suspension was added to 1.5 ml of each of two paints, i.e., Hempasil x3 87500 (Hempel 87509 59151 with Hempasil cross-linker 98950) and Hempasil 77500 (Hempel 77509 30170 with Hempasil cross-linker 97080), and mixed using sterile inoculation needles. The antibacterial activity of the paint-bacterium mix was tested against Vibrio anguillarum strain 90-11-287 (36) in an agar-based well diffusion assay in which V. anguillarum was incorporated into 1.2% agar with 3% (wt/vol) Instant Ocean (IO) salt (Aquarium Systems Inc., Sarrebourg, France), 0.3% (wt/vol) Bacto Casamino Acids (BD, MD), and 0.4% (wt/vol) glucose as previously described (35). Plates were incubated at 25°C for 48 h, whereupon inhibitory activity was detected as a clearing zone in the turbid agar around the wells. Every combination of paint-bacteria and controls was tested in duplicate.

The antifouling effect of each paint-bacterium mix (described above) was tested against Pseudoalteromonas S91, using the same setup as that for testing the antifouling effect of Pseudoalteromonas biofilms on stainless steel over time in sterile seawater. The paint-bacterium mixtures were added to both sides of the coupons and were allowed to dry for 2 h before being placed in microtiter wells. Two hundred microliters of a 1% Sigma sea salt solution was added to each well, containing approximately 10⁶ CFU/ml of Pseudoalteromonas S91, and the number of adhered bacteria was quantified by ultrasonication and plate counts on marine agar after 2 and 48 h. The number of bacteria in the suspension surrounding the coupons was quantified by plate counts without sonication. The bacteria incorporated into the paint (P. piscicida B39bio and P. tunicata J49q-3a) and strain S91 were differentiated based on their pigmentation being yellow, dark purple, and light pink, respectively. Every combination of paint-bacteria and controls was run in triplicate.

Antifouling capacity of Pseudoalteromonas in ship paint submerged in a harbor environment.

Paint-bacterium mixes with Pseudoalteromonas piscicida B39bio and P. tunicata J49q-3a and the Hempasil x3 87500 (Hempel 87509 59151 with Hempasil cross-linker 98950) and Hempasil 77500 (Hempel 77509 30170 with Hempasil cross-linker 97080) paints were prepared as described above. Two controls were included: one in which 1% SSS with no bacteria was added, and one noncoated surface. Every combination of paint-bacteria and controls was tested in duplicate. Acrylic panels with an area of 10 by 20 cm² and a thickness of 0.5 cm were used. The panels were primed using Hempatex high-build 4633 primer from Hempel A/S (a system based on chlorinated rubber binders) to improve the adhesion strength between the panel and the experimental coating. The experimental coatings were applied using an 8-cm doctor blade applicator with a gap size of 400 μm (sheen 1107A/60 014823/I). Panel exposure was done statically on a raft in Jyllinge Harbor, starting April 2012, with the last inspection in December 2012. The salinity was 1.3 to 1.4%, and the seawater temperature increased from 11.3°C in April to 17.6°C in August and fell to 1°C in December; both salinity and temperature were measured in situ with a handheld Professional Plus instrument. Fouling was assessed visually every 4 to 6 weeks, and pictures were taken to document the antifouling performance of each coating.

Statistical analysis.

Comparison of the antifouling capacities of Pseudoalteromonas strains against culturable and total bacterial cell densities on surfaces were done by t test comparisons of log-transformed cell densities (CFU/cm² or cells/cm²).

RESULTS

Survival and antifouling effect of Pseudoalteromonas biofilms on stainless steel over time in sterile seawater.

All Pseudoalteromonas strains grew well in MB, to densities of 10⁸ to 10⁹ CFU/ml, and attached to the steel surface at levels of 6 × 10⁶ to 4 × 10⁷ CFU/cm² (Table 1). The assay had a detection level of 20 CFU (log 1.3) per cm². After 7 days of being submerged in sterile seawater, all strains remained attached to the steel surface, but at a density 100-fold lower than that at the start of the experiment. P. piscicida strains B39bio and A38q-4a and P. tunicata strain J38a-5a were able to survive as biofilms for 53 days in the sterile seawater, but there was a 2.5-log reduction in their numbers over time. The remaining P. tunicata, P. ulvae, and P. antarctica strains were no longer detectable on the steel surface after 53 days of submersion in sterile seawater (Table 1). The four P. tunicata strains survived for 7 days, but none of the strains could be detected after 53 days. They reduced the counts of attaching S91 organisms slightly, but only one (strain J36q-4a) caused a statistically significant reduction. Notably, P. tunicata strain D2, which was included as a presumed positive antifouling control, had little effect in this setup (Table 1).

We then compared the abilities of the eight Pseudoalteromonas strains to prevent attachment of Pseudoalteromonas strain S91. Pseudoalteromonas strain S91 attached at a level of 10⁵ CFU/cm² to noncoated steel surfaces, and attachment was slightly enhanced on surfaces precoated with P. ulvae and P. antarctica (5 × 10⁵ to 6 × 10⁵ CFU/cm²), but the difference was not statistically significant. In contrast, precoating with the two P. piscicida strains significantly reduced S91 attachment after 7 days, to 2.5 × 10² to 1.0 × 10³ CFU/cm². After 53 days, strain S91 could not be detected on surfaces precoated with P. piscicida B39bio, whereas 3 × 10³ CFU/cm² were attached to surfaces precoated with A38q-4a. P. piscicida strain B39bio killed S91 in the seawater suspension (<10 CFU/ml), and strain A38q-4a caused a slight reduction compared to the level on control surfaces (Table 2).

Table 2.

Levels of Pseudoalteromonas strain S91 in sterile seawater in which steel plates coated with potential antifouling Pseudoalteromonas strains were submerged^a

Coating species or control	Strain	No. of Pseudoalteromonas S91 organisms in sterile-filtered water from Jyllinge Harbor (log CFU/ml)
		7 days	53 days
P. piscicida	B39bio	5.0 ± 0.2	<1
	A38q-4a	4.4 ± 0.3	5.0 ± 0.5
P. tunicata	J49q-3a	5.0 ± 1.2	5.7 ± 0.3
	D2	5.5 ± 1.0	5.4 ± 0.2
	J36q-4a	4.3 ± 0.2	6.3 ± 0.2
	J38a-5a	5.8 ± 0.3	6.0 ± 0.2
P. ulvae	H34q-5a	7.4 ± 0.7	6.1 ± 0.2
P. antarctica	E45q-4a	6.2 ± 0.1	6.2 ± 0.0
Controls
Noncoated surface		6.1 ± 0.2	5.3 ± 0.1
MB-coated surface		6.7 ± 0.3	6.1 ± 0.3

^aThe initial level of strain S91 (day 0) was 6 × 10⁵ CFU/ml (log 5.8 CFU/ml).

Enumeration of total bacterial counts by a real-time PCR procedure.

The DNA content in seawater samples extracted with phenol-chloroform was (3.8 ± 0.4) × 10⁻⁷ g/ml. This is slightly higher than that obtained using a commercial kit (NucleoSpin Tissue kit), which yielded (2.1 ± 1.5) × 10⁻⁷ g/ml. Assuming (based on SYBR Gold staining) approximately 10⁷ cells/ml and an average of 2.5 × 10⁻¹⁵ g bacterial DNA per bacterial cell (37), the DNA concentration was 10 times higher than the theoretical value of 2.6 × 10⁻⁸ g/ml, indicating that DNAs from other organisms were also extracted. Also, SYBR Gold staining could underestimate the number of bacteria due to clumps of bacterial cells present or to the amount of DNA/cell differing between the water used in the reference (38) and our coastal water because of different bacterial populations. Comparing C_T values to CFU/ml for the three strains used for standard curves gave the following regressions and correlation coefficients: log CFU/ml = −2.61 × C_T + 35.59 (R² = 0.89) for P. tunicata D2, log CFU/ml = −2.66 × C_T + 37.58 (R² = 0.95) for V. anguillarum 90-11-287, and log CFU/ml = −2.67 × C_T + 37.19 (R² = 0.98) for Phaeobacter inhibens 27-4. The lowest detectable level of bacteria by this procedure was between 1.5 × 10² and 6.8 × 10² CFU/ml. Using a combined standard curve, a water sample contained 6.63 log CFU/ml, which was comparable to the SYBR Gold staining result, which yielded approximately 10⁷ CFU/ml.

Effects on fouling of Pseudoalteromonas strains in natural coastal seawater.

The water used in the fouling batch system was sampled in August 2011 and contained 6.6 ± 0.1 log cells/ml as determined by SYBR Gold staining. The culturable count was 3.0 ± 0.1 log CFU/ml.

After 7 days, the number of cells adhering to both noncoated and MB-coated stainless steel was 10⁵ cells/cm²; however, the MB coating facilitated bacterial adhesion, and after 53 days, 3 × 10⁸ cells/cm² were detected on the MB-coated surfaces, as opposed to 10⁶ cells/cm² on the noncoated surface (Table 3). The levels of bacteria adhering to the surfaces precoated with Pseudoalteromonas were 10⁶ to 10⁷ cells/cm², which was 1 to 2 log units higher than those on the MB-coated surfaces after 7 days (Table 3) (P < 0.001). In particular, P. antarctica facilitated adhesion, and the total adhering cell number was 3 × 10⁸ cells/cm² after 7 days. This higher level was not due to the precoated strains, as the level of Pseudoalteromonas was 10⁴ to 10⁵ CFU/cm², except for on surfaces precoated with P. antarctica, where the level of pseudoalteromonads was 4 × 10⁶ cells/cm². The levels of attaching culturable bacteria were also higher on the Pseudoalteromonas-precoated surfaces than on the control surfaces after 7 days. Based on visual inspection of pigmented colonies on the MA plates, the dominant part of the culturable bacteria was not the precoated Pseudoalteromonas strains, i.e., purple and yellow colonies were in a minority. Based on colony pigmentation, P. piscicida (yellow), P. ulvae (dark green), and P. antarctica (brown) remained at levels of 10³ to 10⁵ CFU/cm² on the steel surfaces. None of the tested P. tunicata strains could be detected by plate counts after 7 days in the system, nor were they detectable when quantified using real-time PCR with primers tagging the alpP gene present in all four P. tunicata strains (30) (Table 3). Macrofouling was observed on all surfaces.

Table 3.

Antifouling effects against different groups of bacteria and antimacrofouling effect of Pseudoalteromonas biofilms on stainless steel after 7 and 53 days of exposure to natural coastal seawater^a

Day	Species or control	Strain	Total no. of bacteriab (cells/cm2)	No. of culturable bacteriac (CFU/cm2)	No. of Pseudoalteromonasb sp. organisms (CFU/cm2)	No. of coated bacteriad (CFU or cells/cm2)	Presence of macrofouling
7	P. piscicida	B39bio	6.8 ± 0.5***	6.3 ± 0.1***	5.2 ± 0.3*	4.3 ± 0.2	+
		A38q-4a	6.9 ± 0.5***	5.9 ± 0.1***	5.8 ± 0.4**	4.0 ± 0.1	+
	P. tunicata	J49q-3a	6.4 ± 0.8***	5.0 ± 0.0***	4.6 ± 0.5	<Dt	+
		D2	6.1 ± 0.3***	4.8 ± 0.2	4.5 ± 0.5	<Dt	+
		J36q-4a	6.7 ± 0.3***	5.6 ± 0.2*	4.6 ± 0.2	<Dt	+
		J38a-5a	6.5 ± 0.4***	5.1 ± 0.1**	4.3 ± 0.3	<Dt	+
	P. ulvae	H34q-5a	6.2 ± 0.4***	5.2 ± 0.2*	3.9 ± 0.3	3.2 ± 0.3	+
	P. antarctica	E45q-4a	8.5 ± 0.5***	7.2 ± 0.2**	6.6 ± 0.3**	5.0 ± 0.5	++
	Noncoated surface control		5.1 ± 0.4	4.0 ± 0.0	3.6 ± 0.3		+
	MB-coated surface control		5.1 ± 0.3	4.8 ± 0.0	4.3 ± 0.4		++
53	P. piscicida	B39bio	6.9 ± 0.3***	5.3 ± 0.1**	4.0 ± 0.5*	<Dt	++
		A38q-4a	7.8 ± 0.3**	6.0 ± 0.1	5.4 ± 0.5***	<Dt	++
	P. tunicata	J49q-3a	7.7 ± 0.5	4.6 ± 0.1**	6.1 ± 0.4**	<Dt	++
		D2	8.0 ± 0.3*	6.1 ± 0.4	6.3 ± 0.1**	<Dt	++
		J36q-4a	5.7 ± 1.3**	5.9 ± 0.6	5.4 ± 0.5**	<Dt	++
		J38a-5a	6.9 ± 0.5***	4.9 ± 0.0**	5.8 ± 0.5***	<Dt	++
	P. ulvae	H34q-5a	7.0 ± 0.3***	5.3 ± 0.3*	4.7 ± 0.3***	<Dt	++
	P. antarctica	E45q-4a	8.0 ± 0.3**	6.0 ± 0.1	5.9 ± 0.3***	<Dt	+++
	Noncoated surface control		6.0 ± 0.7	5.3 ± 0.1	3.3 ± 0.1		++
	MB-coated surface control		8.5 ± 0.3	6.6 ± 0.4	2.4 ± 0.4		+++

^aThe level of macrofouling was visually judged as follows: +, low level; ++, high level; and +++, very high level. Counts on Pseudoalteromonas-precoated surfaces were compared to those on MB-coated surfaces by using the t test. *, P < 0.05; **, P < 0.01; ***, P < 0.001.

^bThe number of bacteria was quantified using real-time quantitative PCR.

^cQuantified by plate counts on 50% marine agar.

^dQuantified by counting pigmented colonies after plate counts on 50% marine agar and by quantitating the AlpP gene of P. tunicata by real-time quantitative PCR. Dt, detection level.

After 53 days, none of the precoated Pseudoalteromonas organisms could be detected on the surface, based on the appearance of pigmented colonies on MA plates. Levels of Pseudoalteromonas varied from 3 × 10² cells/cm² to 2 × 10⁶ cells/cm² when quantified using Pseudoalteromonas-specific primers. The lowest level was found on the steel plates not precoated with pseudoalteromonads, and the levels on Pseudoalteromonas-precoated plates were statistically higher than those on the MB-coated plates (Table 3). A dense fouling was observed by SYBR Gold staining on all surfaces, independent of precoating (noncoated, MB coated, or coated with our Pseudoalteromonas strains) (see Fig. S1 in the supplemental material). The fouling consisted mainly of bacteria (Fig. S1), with the MB-coated surface attracting the densest layer (Fig. S1).

In vitro antibacterial effect of Pseudoalteromonas incorporated into ship paint.

Suspensions of P. tunicata and P. piscicida were prepared from lawn-grown bacteria and contained 3 × 10⁹ CFU/ml. These suspensions were mixed into the paints and tested in a well diffusion assay. However, all samples, including the non-Pseudoalteromonas paints, caused inhibition zones in the Vibrio anguillarum layer, likely due to the organic solvents in the paint. When painted steel coupons were submerged in suspensions of strain S91 in sterile seawater, approximately 10³ to 10⁴ CFU/cm² adhered after 48 h, independent of the addition of pseudoalteromonads to the paint.

Antifouling capacity of Pseudoalteromonas in ship paint in vivo.

The amount of fouling accumulated on the panels was visually inspected 6 times during the 7-month submersion in Jyllinge Harbor. Very little fouling was seen during the first 4 months, and no difference was noted between the panels with bacteria embedded in the paint and the regular painted panels. However, in October (approximately 10°C), after 5 1/2 months, the panels with bacteria embedded were less fouled than the control panels (Fig. 1). This difference was seen for only one of the two paints, the Hempasil 77500 mixture. When we inspected the panels 1 1/2 months later, the Hempasil 77500 mixture with bacteria was slightly less fouled than the control panels.

Fig 1.

Fouling development on paints with incorporated P. piscicida strain B39bio (middle two panels) and P. tunicata strain J49q-3a (leftmost two panels) after 166 days in Jyllinge Harbor. Controls (rightmost two panels) included the paints with 1% Sigma sea salt added. The two series, each with three panels, are duplicates and were placed at separate locations in the submerging rack.

DISCUSSION

Marine bacteria interact with pro- and eukaryotic organisms, and several studies have described how bacteria in aggregates or microbial biofilms can act as attractants for other bacteria (38) or eukaryotic organisms (39, 40). In contrast, they also act as repellants against both other bacteria (21, 41) and a range of eukaryotic organisms (42–44). The present study in combination with previous findings (30) demonstrates that depending on the model organisms, marine bacteria may either repel (e.g., Pseudoalteromonas S91 and Ulva spores) or attract (e.g., the natural microbiota of seawater) other bacteria. Also, time influences the outcome, as several pseudoalteromonads acted as attractants for the natural seawater prokaryotic population for the first week of exposure but had a repelling effect after 1 1/2 months. Also, a possible repelling effect was seen when panels were painted with Pseudoalteromonas-containing paint and left in a harbor area.

The Pseudoalteromonas strains used in this study were selected due to their antibiotic activity in in vitro model systems and were found to have antifouling activity against bacteria and algal spores (30), and it is tempting to hypothesize that the antibiotic activity is a component of the antifouling activity in monoculture model systems. However, Long et al. (45) demonstrated that an alteromonad antibiotic, 2-n-pentyl-4-quinolinol, reduced attachment of some phylotypes to surfaces but enhanced attachment of other phylotypes. Similarly, Grossart et al. (38) found that the production of antibiotic by a biofilm community did not affect newcomers to the biofilm. Of the four Pseudoalteromonas species in the present study, three (P. piscicida, P. tunicata, and P. ulvae) represent the pigmented, antibiotic-producing subgroup, whereas P. antarctica, despite a brownish colony color, represents the nonpigmented, non-antibiotic-producing cluster (46, 47). In the simple (sterile seawater) systems, we observed no consistent difference in repelling or attracting characteristics between P. antarctica and the pigmented strains, further supporting the above-mentioned studies that could not correlate antibiotic production to a subsequent repelling effect (38, 45). Holmström et al. (24) compared the antibacterial, antifungal, and antilarval activities of 10 different Pseudoalteromonas strains, including P. piscicida. Only P. tunicata was active in all bioassays, but we found in both our previous study (30) and the present study that P. piscicida was a stronger antagonist. This difference could be strain related, as the P. piscicida strain included in the study of Holmström et al. (24) had only a limited antibacterial effect in agar assays, whereas our strains caused large clearing zones in agar antagonism assays. The P. piscicida strains in the present study reduced Pseudoalteromonas S91 adhesion after 7 days, and one of the P. piscicida strains was the only pseudoalteromonad strain reducing S91 adhesion after 53 days (Table 1). This antifouling effect was likely caused by the bactericidal activity of the strain (Table 2). P. piscicida strains also appeared to be the most firmly adhered strains, remaining in the biofilm at a level of 10⁴ to 10⁵ CFU/cm² after 53 days. While the bioactive compounds of P. tunicata have been studied in detail (27), much less is known about the chemical structure of P. piscicida bioactive compounds. A high-M_w protein has been identified as an antibacterial compound (48), and an alkaloid (norharman) was identified as a cytotoxic compound (49). P. piscicida also produces brominated peptides (46, 47) that have a cytotoxic effect (50). Despite several attempts, we have not been able to purify the antibacterial compound(s). Interestingly, the two P. piscicida strains did promote adhesion of natural pseudoalteromonads, albeit not to the same degree as the nonantibacterial P. antarctica strain and two of the P. tunicata strains (Table 2, data for 53 days).

The four P. tunicata strains all formed biofilms, but these were not stable when exposed to natural seawater or to Pseudoalteromonas strain S91 (Table 1). P. piscicida, P. ulvae, and P. antarctica could all be detected after 7 days on steel surfaces submerged in natural seawater, but P. tunicata was not detectable by either counts (pigment) or alpP gene detection (Table 3). Rao et al. found that P. tunicata easily outcompeted other bacteria in mixed biofilms, expect for Phaeobacter (Roseobacter) gallaeciensis, which eradicated P. tunicata in a biofilm flow cell system (3) and in axenic Ulva australis systems (51). We have not determined the composition of the natural bacterial biofilm developing on top of the Pseudoalteromonas biofilm (Table 3), but we used water from Jyllinge Harbor, where we repeatedly detected Phaeobacter species on biotic and abiotic surfaces in the summer months of 2009 to 2012 (L. Gram, B. Barker Rasmussen, N. Bernbom, Y. Y. Ng, C. H. Porsby, and T. Brinkhoff, unpublished data), and this could potentially explain the disappearance of P. tunicata from the biofilm (3). Also, the disappearance of P. tunicata could be caused by the AlpP protein, which generates H₂O₂ and mediates biofilm cell death (28).

P. tunicata strain D2 was included as a presumed positive antifouling control, but this strain did not perform as well as, e.g., P. piscicida, in our setup. Holmström et al. (24) cultured bacteria by using Väätänen nine-salt solution (VNSS) medium, which contains glucose, whereas we used MB, which does not contain glucose and has a markedly higher concentration of iron. Bacterial secondary metabolism is very dependent on growth substrates, and one could speculate that D2 does not produce the same array of extracellular bioactive compounds in MB as in VNSS. The present study used a setup quite different from those of previous experiments with strain D2, and we suggest that the longer exposure, which allowed AlpP-mediated autolysis to take effect, and the use of natural seawater that probably harbors AlpP-resistant strains could explain the low antifouling effect of D2.

P. piscicida had antifouling activity against Pseudoalteromonas S91 in a sterile seawater system (Table 1), but in mesocosms with natural seawater, all strains appeared to have an initial attractant effect but a slight repelling effect over a longer time (Table 3). Many microorganisms facilitate growth of other microorganisms, and cultivation of marine microorganisms can be enhanced significantly by the presence of coexisting bacteria (52, 53). Also, many marine bacteria display chemotactic behavior (54) and are attracted to other bacteria (54, 55). The slight repellant effect over a longer time (Table 2) was observed compared to the MB-coated plate. However, because of nutrient richness, this control plate may have attracted (de novo) a more stable natural community than could be established on plates coated with monocultures of pseudoalteromonads. It should be noted that the level of Pseudoalteromonas after 53 days was higher on all Pseudoalteromonas-precoated plates, indicating a preference for attracting the same genus.

The vast majority of studies testing antifouling agents have used pure compounds (biocides) or extracts as paint additives (4, 56). Since we have not been able to extract a stable antibacterial compound from P. piscicida, we did not, as others have done (13), use bacterial extracts but used a thick bacterial suspension as an additive to the paint. The bacteria were killed immediately when mixed into the paint, and our system is hence different from other approaches where live bacteria are embedded in matrices of gel or carrageenan (11, 57). Both a P. tunicata and a P. piscicida strain in one of the two paints diminished fouling compared to that on the control panels, and further trials in warmer waters are in progress. Both strains produce a range of bioactive secondary metabolites (27, 46), and we speculate that these were extracted into the paint (by organic solvents) and acted as chemical antifouling compounds that slowly leaked from the paint. The lack of activity of the strains in the second paint could be due to the reference plate performing extremely well or to less binding of bacterial secondary metabolites in the paint, resulting in diffusion of these into the water.

Estimation of bacterial densities is a key analysis in estimating repellant or attractant effects. While this is simple in defined systems with known organisms (30), it is more complicated in systems with high levels of unculturable organisms. We chose to use universal primers following the procedure of Einen et al. (33). In their study (33), biomass was estimated based on the number of 16S rRNA genes in a bacterium (Escherichia coli) and an archeal species (Archaeoglobus fulgidus). We decided to use a standard curve comparing C_T with CFU/ml based on the combined standard curves for three species of marine bacteria commonly found as aggregate and fouling communities (51, 58, 59): a Roseobacter-clade strain, a Pseudoalteromonas strain, and a Vibrio species. When comparing estimated cell counts in water samples by using the PCR-based method and SYBR Gold staining, we found excellent agreement, as did Einen et al. (33).

Our study demonstrates the need for a range of in vitro and in vivo assays for developing antifouling strategies. The convincing in vitro results (30) did not translate into short-term effects in seawater mesocosms, but both mesocosm and paint studies indicated an antifouling effect on a longer time scale and emphasized the need for systems closer to real-life scenarios.