Monospecific Biofilms of Pseudoalteromonas Promote Larval Settlement and Metamorphosis of Mytilus coruscus

As a stage of life cycle, larval settlement and metamorphosis are critical processes for persistence of many marine invertebrate populations. Bacterial biofilms (BFs) could induce larval settlement and metamorphosis. Pseudoalteromonas, a widely distributed genus of marine bacteria, showed inductive effects on several invertebrates. However, how Pseudoalteromonas BFs induce settlement and metamorphosis of Mytilus coruscus remains unclear. Pseudoalteromonas marina BFs with the highest inducing activity were further investigated to define inductive cues. Surface-bound products of P. marina BFs could induce larval settlement and metamorphosis. P. marina BFs treated with formalin, antibiotics, ultraviolet irradiation, heat and ethanol significantly reduced inductive effects and cell survival rates. The confocal laser scanning microscopy and the biovolume analysis showed the dominance of α-polysaccharides on P. marina BFs. Treatment of BFs with amylases, proteases and lipase led to the decrease of inducing activity, suggesting that inductive cues of P. marina BFs may comprise of molecular domains of polysaccharides, proteins, and lipids. Finding inductive cues of BFs could put forward further studies about the mechanism of larval settlement and metamorphosis of marine invertebrates.

water-borne and surface-bound cues are need for Shewanella sp. 1 BFs to induce larval settlement and metamorphosis 14 . However, how EPS and its components affected inductiveness of Pseudoalteromonas BFs remain unclear.
M. coruscus is distributed at the temperate zone of coastal waters in East Asia. It is a common economic shellfish and a major fouling organism in China 30 . Thus, investigation of M. coruscus settlement has a great ecological and industrial significance 30 .
In the present study, inducing activities of seven Pseudoalteromonas strains were tested. To elucidate inductive cues, P. marina, which had the highest inducing activity was used as a model strain. In this study we investigated: (1) larval responses to varying species of Pseudoalteromonas in order to examine whether Pseudoalteromonas BFs showed the similar inducing or inhibiting activity to the mussel M. coruscus; (2) whether the surface-bound products of BFs have inducing activity; (3) the inducing activity and cell mortality of BFs treated with formalin, ethanol, heat and ultraviolet irradiation, in order to determine properties of BF inductive cues; and (4) the inductive effect of BFs treated by amylases, proteases, and a lipase to determine bioactive moieties of inductive cues.
Inductive effects of surface-bound products. The surface-bound products of P. marina BFs significantly promoted larval settlement and metamorphosis at all tested concentrations (χ 2 : df = 4, p < 0.001, Fig. 3). The inducing activities of surface-bound products at concentrations of 0.5 mg l −1 and 5 mg l −1 were similar to that of P. marina BF. The surface-bound products at the concentration 50 mg l −1 were less inductive than P. marina BF (Fig. 3). After 24 h, 28% and 25% larvae were swimming in the presence of 0.5 mg l −1 of surface-bound products (Fig. 4A) and BFs (Fig. 4B), while more than 50% of larvae were swimming in autoclaved filtered seawater (AFSW, Fig. 4C). After 48 h, 28% post-larvae were observed in the presence of the 0.5 mg l −1 of surface-bound products Inducing activity of treated P. marina biofilms. The inducing activity of P. marina BFs decreased significantly (χ 2 : df = 10, p < 0.001) after treatment with formalin, antibiotic, ultraviolet irradiation, heat and ethanol (Fig. 5A). Similarly, cells survival significantly (χ 2 : df = 10, p < 0.001, Fig. 5B) decreased. The FF treatment led to low inducing activity and 100% cell mortality in P. marina BFs (Fig. 5). The antibiotic treatment significantly changed cell survival and inductiveness of BFs (p < 0.05) (Fig. 5). No cell survived after the treatment with the highest treatments of 100E and 100 H (Fig. 5B).

Discussion
Pseudoalteromonas gained attention due to its ecological significance 18,23 . Bacteria belonging to Pseudoalteromonas genus have been commonly observed in marine BFs 17 . Pseudoalteromonas species induce larval settlement of the polychaete Hydroides elegans 19 and the mussel M. coruscus 14 and prevent settlement of bacteria and fungi, algal spores and invertebrate larvae 23,24 . In present study, BFs of seven different Pseudoalteromonas species could induce larval settlement and metamorphosis of M. coruscus.
The inducing activity of Pseudoalteromonas species isolated from natural BFs and the gut of the mussel showed similar inducing activities. This indicates that there is no relation between bacterial origin and inducing activity of settlement. Similar results were observed for other invertebrate species 19,21 .    Our study showed that surface-bound products of P. marina BFs could trigger the similar percentages of crawling behavior of M. coruscus as BFs of this bacterium. In contrast, most of larvae keep swimming in AFSW. Larval response to inductive cues with crawling behavior was also shown for other species such as M. galloprovincialis 36,37 . Similarly, the polychaete larvae of H. elegans stop swimming and start crawling within a few minutes of exposure to inductive bacterial BFs 5 . In addition, the crude extracts of surface-bound products of Alteromonas sp. 1 BFs could not induce larval metamorphosis of M. galloprovincialis 37 . Surface-bound products of P. marina BFs could induce larval metamorphosis of M. coruscus. This difference between M. galloprovincialis and M. coruscus might be due to the different species or composition of surface-bound products. Alteromonas sp. 1 BFs scraped off from Petri dishes were sonicated for 5 min and the crude extracts of surface-bound products contained both surface-bound products and bacterial cellular contents 37 . In present study, P. marina BFs scraped off from Petri dishes were mixed with PBS by pipetting instead of sonication and crude extracts of surface-bound products  www.nature.com/scientificreports www.nature.com/scientificreports/ had less cellular contents. On the other hand, surface-bound products in the present study contained only compounds >3.5KDa, while inductive cues in other studies had different molecular weights, such us, 1-10 KDa (scyphozoan) 15 , ≤0.3 KDa (oyster) 38 and ≤3 KDa (M. galloprovincialis) 37 .
It is shown that water-borne cues could induce larval metamorphosis 29 of Phestilla 39 , Adalaria proxima 40 , and Balanus amphitrite 41 as well as larval settlement of oysters 42,43 . For Mytilus species, like M. coruscus 14 and M. galloprovincialis 37 , water-borne cues could induce larval settlement but not metamorphosis. Similarly, physical properties of natural BFs, instead of water-borne cues, induced settlement of a pearl oyster 44 . Thus, in the present study, we investigated the effect of surface-bound cues.
Previous studies have shown that biofilms of bacteria isolated from surface of adult B. amphitrite 41 and M. coruscus 14 could induce larvae metamorphosis. In the present study, bacteria carried by larvae showed no inductive effect on larval settlement and metamorphosis, it may be due to that low numbers of bacteria could not form an effective biofilm. Bacterial matrix could promote biofilm formation 45 . Thus, in the presence of surface-bound products, biofilms being seeded with bacteria from larvae might form easier and show inducing activity. Thus, the inducing activities of biofilms seeded by larvae need to be examined in the future.
To investigate characteristics of inductive cues, P. marina BFs were treated by antibiotics, formalin, ethanol, heat, and UV. All treatments led to the decrease of the inductive activity of the P. marina BFs and the increase of the mortality of bacteria. Our results were consistent with previous research with larvae of M. galloprovincialis 33,37 and M. coruscus 14 . Our and previous studies 14,33,37 suggested that inductive cues were susceptible for above-mentioned treatments, and the viability of bacterial cells was important for the production of inductive cues. Similar results were obtained in experiments with treated BFs and other larvae of marine invertebrates [46][47][48] . For example, formalin-killed BF of Pseudomonas marina induced low settlement and metamorphosis (less than 9%) of Janua 46 . Natural BFs treated with ethanol (5% and 10%), heat (40 °C) led to the reduction of inducing activity of larval metamorphosis of sea urchins 47 . Antibiotic treatment reduced the percentages of larvae settled on Roseobacter and Proteobacteria BFs 48 .
Analysis of CLSM images showed that P. marina BFs had the higher biovolume of the α-polysaccharides than those of proteins, β-polysaccharides, and lipids. In opposite, the biovolume of α-polysaccharides was lower than those of β-polysaccharides and proteins in BFs of Salmonella enterica serotype Agona 49 . This suggested that components of BFs varied between bacterial species.
In order to investigate the inductive cue from P. marina further, BFs were treated with different enzymes. Treatment of BFs with α-amylase reduced its inductiveness. This suggested that exopolysaccharides or glycoproteins are involved in larval metamorphosis of M. coruscus. Exopolysaccharides or glycoproteins of BFs bound to lectins were involved in larval settlement and metamorphosis of M. galloprovincialis 37 and Janua (Dexiospira) brasiliensis 50 . Glycoproteins were involved in larval settlement of marine invertebrates. For example, the settlement-inducing protein complex, a glycoprotein consisting of lentil lectin-binding sugar chains, induced larval settlement of B. amphitrite 51,52 . A tri-peptide with arginine at the C-terminal was inductive for Crassostrea virginica 53 , and glycoproteins in the organic matrix of pacific oyster shells were inductive for Crassostrea gigas settlement 54 . A chemical identity of bacterial exopolysaccharide or glycoprotein that induced M. coruscus larval settlement requires further investigations.
Moreover, proteases (trypsin and papain) reduced inductive activity of P. marina BFs, suggesting that the bioactive moieties of inductive cues were proteinaceous. Earlier studies have proved that proteins could induce larval settlement or metamorphosis of the oyster 54 , jellyfish 55 and barnacles 56 . However, purified proteins inducing larval settlement and metamorphosis of M. coruscus are still unknown.
The result of lipase treatment suggests that lipids may be involved in larval settlement and metamorphosis. In previous studies, fatty acids were suggested as chemical cues for larval settlement or metamorphosis. For example, free fatty acids isolated from a coralline red alga induced settlement and metamorphosis of Pseudocentrotus depressus and Anthocidaris crassispina 57 . Free fatty acids isolated from sand matrix of adult tubes of polychaetes induced larval settlement and metamorphosis of Phragmatopoma species 58,59 . For barnacle larvae, lipids played important roles in permanent adhesive by creating a supportive environment and reducing bacterial biodegradation of nascent adhesive plaque 60 . The role of lipids in settlement and metamorphosis of M. coruscus still needs to be researched.
In summary, all tested Pseudoalteromonas species could promote M. coruscus larval settlement and metamorphosis. Surface-bound products of biofilms contained inductive cues. The inductive cues were susceptible to formalin, ethanol, heat and ultraviolet irradiation treatments. The viability of bacterial cells was important for the production of inductive cues. The inductive cues induced larval settlement and metamorphosis may include polysaccharides, proteins and lipids.

Materials and Methods
Bacterial isolation and cultivation. Bacteria were isolated from four weeks old natural BFs and gut of the wild two-year adult mussel M. coruscus. Suspensions of natural BFs were prepared as described in the previous study 19 . Briefly, natural BFs which formed on clean glass at seawater of China East Sea (122°46′ E; 30°43′ N), were scraped off from the glass slides and suspended into autoclaved filtered seawater (AFSW). The adult mussels were collected from the same area, washed three times by AFSW and dissected on ice by sterilized tools. The gut tissues were pounded and suspended in AFSW. The suspensions of natural BFs and gut cells were carried out 10-fold serial dilutions to the dilution of 10 4 times with AFSW. Individual bacteria were separated by spreading diluted cell suspension on Zobell 2216E agar plates. One liter of Zobell 2216E agar plates contained 5.0 g peptone, 1.0 g yeast extract, 0.01 g Ferric citrate and 15.0 g agar. All reagents were purchased from Sigma Company (United States). Individual bacterial colonies were selected based on their shape, morphology and colour. Individual colonies were purified three times through plate streaking. Prior to the experiments, all bacterial strains were stored at −80 °C with cryoprotectant (0.9% physiological saline, 15% glycerol). (2020) 10:2577 | https://doi.org/10.1038/s41598-020-59506-1 www.nature.com/scientificreports www.nature.com/scientificreports/ Identification of Pseudoalteromonas. Prior to identification, bacterial strains were cultured in Zobell 2216E medium for 16-24 h. After centrifugation, broth culture was discarded and cells were collected for DNA extraction. TIANamp Bacteria DNA Kit was used for genomic DNA extraction following manufactures protocol (TIANGEN Biotech Company, Beijing, China). The primers 27 F ((5′-AGA GTT TGA TCC TGGCTC AG-3′) and 1492 R (5′-GGT TAC CTT GTT ACGACT T-3′) were used to specifically amplify the 16 S rRNA gene sequence following a modified PCR procedure 61 . The reaction mixture of the amplification contained 0.5 μl of 27 F (10 μM), 0.5 μl of 1492 R (10 μM), 0.5 μl of DNA (100 ng μl −1 ), 0.3 μl of Taq (5 U μl −1 ), 2 μl of MgCl2 (25 mM), 0.5 μl of deoxynucleotide triphosphates (10 mM each), and 20.5 μl of sterilized distilled water. The amplification was performed using the following PCR conditions: an initial denaturing step (94 °C, 5 min), 35 cycles of denaturation (94 °C, 30 s), an annealing (55 °C, 30 s), and an elongation (72 °C, 2 min). An Eppendorf Mastercycler (Eppendorf, Germany) was used. The PCR products were sent to Sangon Biotech Company (Shanghai, China) to get 16 S rRNA gene sequences. These sequences were blasted using the National Center for Biotechnology Information (NCBI) website to identify close matches.
Preparation of biofilms. The monospecific BFs of different Pseudoalteromonas strains were prepared following modified published method 19 . Briefly, Pseudoalteromonas strain was cultured in 80 ml of Zobell 2216E medium at 25 °C for 24 h with shaking (200 rpm). The bacterial cells were harvested from culture by centrifugation (1600 g, 15 min). Cells were re-suspended in AFSW and washed thrice to remove broth residues. Finally, cell pellets were re-suspended in AFSW again as a bacterial stock solution. One ml diluted bacterial stock solution (1:1000) was stained for 5 min with acridine orange solution (AO, 0.1%) and then transferred to a 0.22 μm Nuclepore Track-Etch Membrane (Whatman, UK) by negative-pressure filtration (diameter of filter mesh: 1.0 cm). The number of bacteria on a membrane was counted in 20 randomly selected fields of view under 1000 × magnification of Olympus BX51 (wavelengths of excitation/emission (Ex/Em) = 490/519 nm) 62 . Bacterial density was re-calculated to cells ml −1 using Eq. (1).
where S was concentration of stock solution (cells ml −1 ); C was cells of field of view (cells cm −2 ); D was diameter of filter mesh; V was 1 ml which the volume of diluted bacterial stock solution transferred to membrane.
The stock solution of each tested Pseudoalteromonas strain was diluted with AFSW to prepare four initial concentrations of bacterial suspension (1 × 10 6 , 1 × 10 7 , 1 × 10 8 and 5 × 10 8 cells ml −1 ). Twenty milliliters of bacterial suspension and a microscopic glass slide (2.5 cm × 3.8 cm) was added to a sterilized glass Petri dish (ASONE Corporation, Japan). The Petri dishes were incubated at 18 °C for 48 h until BF was developed. Unattached bacteria were removed by washing the dishes with AFSW three times. The monospecific bacterial BF was the film remained on a glass slide.

Detection of bacterial density of biofilms.
BFs were soaked at 5% formalin solution for 48 h. Fixed BFs were washed with AFSW thrice to remove residual formalin and stained by AO (0.1%, v/v) for 5 min. As mentioned above, bacterial density (cells cm −2 ) was counted using an Olympus BX51 microscope at 1000 × magnification and 10 fields of views. The assay had three independent replications. Fertilization and larval culture. Two years old of M. coruscus were collected from Shengsi, Zhoushan, China (122°44′ E; 30°73′ N). Fertilization was performed based on the methods of Yang et al. 63,64 . Clean mussels were incubated on ice overnight. Then, mussels were put into 10 liters polycarbonate tanks. Spawning was stimulated by addition of ca 21 °C filtrated seawater (FSW). The spawning mussel was transferred to a new 2 l glass beaker-containing FSW. The available generative cells were collected from the glass beakers for fertilization. Sperm and eggs were incubated for 20 minutes to fertilize them. A nylon plankton net (mesh size: 20 μm) was used to collect and wash fertilized eggs to remove excess of sperm. Fertilized eggs were incubated at 18 °C steadily at dark. After two days of incubation without feeding, a swimming straight-hinge veliger larval stage was obtained. Veliger larvae were cultured at glass beakers at density of 5 larvae ml −1 . Each larva was fed with a diet of Isochrysis zhanjiangensis at 1 × 10 4 cells every day. Fresh FSW was used to change culture water every two days, until the larvae developed into a pediveliger stage. This stage was used for the settlement and metamorphosis bioassays (see below).
Settlement and metamorphosis bioassays. The settlement and metamorphosis bioassay was performed in glass Petri dishes (Ø 64 × 19 mm) (ASONE Corporation, Japan) according to the previous publications 19 . Each dish contained twenty competent pediveliger larvae, 20 ml AFSW and a developed BF (see above). Petri dishes were incubated at 18 °C without light for 48 h, and larvae were observed using a stereoscopic microscope (Olympus, Japan) at 1000 × magnification. The number of settled and metamorphosed larvae was counted and later expressed as percentages. "Settled larvae" stop swimming in water, settle and crawl on the BFs. "Post-larvae" changed both of larval behavior and the body restructure, including the settlement of larvae, the losing of velum, the development of gills, and the produce of juvenile/adult shell-dissoconch. Alive larvae that were lying on the bottom and not moving were defined as "Lying". In opposite if larvae were crawling on the bottom they were group into the "Crawling" category. If larvae were not responding to the gentle needle touch and only had transparent shells, they were defined as "Dead". A sterilized, clean, BF-free glass slides were placed in Petri dishes and used as negative controls. Each treatment contained nine replicates.
The collection of surface-bound products. The  www.nature.com/scientificreports www.nature.com/scientificreports/ light. Surface-bound products of P. marina BFs were collected followed a modified approach of Bao et al. 37 . BFs were scraped from 800 dishes (the total area is 2.3 × 10 4 cm 2 ) and suspended in 1 × PBS (Phosphate Buffer Saline, pH = 7.4). After centrifugation (12,000 g, 20 min), the precipitate was discarded, and the supernatant was filtered using 0.22 μm-pore size filter (Sangon Biotech Co., Ltd, Shanghai, China) to remove bacterial cells. The filtrate containing surface-bound products was transferred to 3.5 KDa cut off dialysis bags (MYM Biological Technology Company Limited, USA) to remove salts. The dialysis bag was incubated for 24 h at the room temperature. Surface-bound products were collected from the bag and were freeze-dried.
Different concentrations (0.5, 5, and 50 mg l −1 ) of surface-bound products were prepared by mixing with AFSW. Each Petri dish (Ø 64 × 19 mm) (ASONE Corporation, Japan) contained 20 milliliters of surface-bound products solution and twenty pediveliger larvae. Larvae were observed using a stereoscopic microscope (Olympus, Japan) at 1000 × magnification as mentioned above in the settlement and metamorphosis bioassay. Petri dishes containing AFSW (without BF and extracellular products) were served as a negative control. P. marina BFs were served as positive controls. Each tested concentration and control group set had six replicates. The percentages of post-larvae and larvae in different behavior group were recorded after 24 h, 48 h, 72 h, and 96 h.
Scanning electron microscope (SEM). P. marina BFs formed at the initial concentration of 1 × 10 7 cells ml −1 for 48 h. BF was washed three times with AFSW and fixed overnight by 5% formaldehyde solution. Fixed BFs were washed by 1 × PBS solution three times to remove residual formalin. BFs were dehydrated by alcohol series and then dried at a room temperature. Finally, BFs were gold coated and visualized with a Hitachi S-3400 field emission SEM at 4 KV of accelerating voltage and magnification of 6500 × and 2700 × .
Biofilms stained with fluorescent dyes. BFs were stained following the method of González-Machado et al. 49 . Extracellular products of BFs were stained by fluorescent dyes and observed using confocal laser scanning microscopy (CLSM). The following dyes were used: the concanavalin A, tetramethylrhodamine conjugate (ConA-TMR) (Invitrogen Company, Germany) for α-polysaccharides at Ex/Em = 552 nm/578 nm, calcofluor white M2R (CFW) (Sigma Company, United States) for β-polysaccharides at Ex/Em = 254 nm/432 nm, fluorescein isothiocyanate isomer I (FITC) (Sigma Company, United States) for proteins Ex/Em = 495 nm/519 nm, and DiIC18 (5)  Treatment of P. marina biofilms. P. marina BFs were treated with different agents according to previous studies 14 . Formalin-treated BFs were prepared by immersing BF in 5% formalin solution for 24 h. The UV-treatment BF were prepared by exposure under 5 W m −2 , 254 nm of UV radiation for 1 h. The ethanol-treatment was prepared by immersing BFs in 10% or 100% ethanol solution for 0.5 h. The heat-treatment was prepared by heating BFs at 40 °C or 100 °C for 0.5 h. The antibiotic-treatment was prepared by immersing BFs in different concentrations of antibiotic solutions (0.1 A, 1 A, and 10 A) for 2 h. The 0.1 A antibiotic solution was a mixture of the penicillin G potassium (0.1 mg ml −1 ) and streptomycin sulfate (0.1 mg ml −1 ), 1 A and 10 A were ten-fold and hundred-fold concentrated solutions of 0.1 A. In order to remove remaining chemicals (formalin, ethanol, and antibiotics), treated BFs were washed nine times using 20 ml of AFSW. After the treatment of UV and heat, BFs were rinsed with 20 ml of AFSW for three times.
To investigate the bioactive moiety of the inductive cues of P. marina BFs, different enzymes treatment were performed following previously published method 65,66 . The enzymes' stock solutions were prepared using mixture of the Tris-HCl (10 mM, pH = 7) and 50% glycerol. Working solutions were diluted with AFSW. The following concentrations of enzymes were used: trypsin (from bovine pancreas) − 0.01 mg ml -1 , papain (from papaya latex) − 0.2 mg ml -1 , β-glucuronidase (from bovine liver) − 0.01 mg ml -1 , sulfatase (from Patella vulgata) − 0.5 mg ml -1 , and α-amylase (from Bacillus sp.) − 0.01 mg ml -1 . The lipase (from Thermomyces lanuginosus) was diluted using Tris-HCl (10 mM, pH = 7) to working concentration of 5 μg ml -1 . All enzymes were obtained from Sigma (USA). The enzyme-treatment was incubating BF at enzyme solution at 25 °C for 3 h. BFs were washed with AFSW three times to remove extra enzyme and then were used to larval settlement and metamorphosis bioassays (see above).

Rates of survive and dead cell of treated biofilms.
Rates of living and dead cells of BFs were examined following previously published methods 14 . BFs were stained by the 5-cyano-2, 3-ditolyl tetrazolium chloride (4 mM) and the 4′, 6-diamidino-2-phenylindole (1 μg l −1 ). Dead cells and living cells were observed by Olympus BX51 epifluorescence microscope under excitation/emission = 365/455 nm and 1000 × magnification. Ten random views of three different BFs were used to calculate survival rate of cells of BFs.