Microbial colonization of benthic environments in Antarctica: responses of microbial abundances, diversity, and activity and larval settlement to natural or anthropogenic disturbances and search for secondary metabolites

In aquatic environments, the formation of microbial biofilms (biofouling)is the first step in the settlement and colonization of substrates by benthic organisms. Being "hot-spots" of biodiversity, microbial biofilms are expected to be a source of secondary metabolites with biotechnological potential. The ANT-Biofilm project was aimed at assessing for the first time in Terra Nova Bay (Ross Sea, Antarctica) the structure and function of biofilm communities (microbes and macrobenthos) and their response to environmental and/or anthropogenic forcings in extreme environments. An interesting part of the project involved the characterization of strains of bacteria and actinomycetes isolated from the biofilm in order to evaluate their biotechnological potential as sources of new bioactive molecules.

Objective 1 Physicochemical and microbiological analyses on water and biofilm (by UO 1- CNR and UO 2- UniMessina).

Objective 2 Possible interrelationships between microbial biofilm and macrobentos colonization (by UO3 UniGenova).

Objective 3 To explore the ability of microbial isolates to produce biofilms and secondary metabolites and to secrete hydrolytic enzymes (by UO 4 UniInsubria)

Objective 4 Characterization of study sites for physicochemical, trophic and microbiological aspects (by all Operational Units)

UO1 was in charge of coordinating field activities, together with UO3, which oversaw the placement of colonization structures mounting PVC and polyethylene (PE) panels, thanks to ENEA support. For Objective 4, measurements of temperature, salinity, fluorescence, and pH were conducted by multiparameter probe; nutrient content, abundance of fecal contamination indicator bacteria, total prokaryotic abundance, its viable fraction, and heterotrophic culturable fraction were estimated. Research activities related to Objective 3 were conducted through isolation of bacterial strains from the biofilm developed on PVC panels submerged in Nov 2017 at -5 m at RB and PTS.

Objectives 1 and 4 Activities related to Objective 1 involved physicochemical and microbiological analyses of water and biofilms (UO 1- CNR and UO 2- UniMessina). UO1 was also in charge of coordinating field activities, together with UO3, which took care of the placement of colonization structures thanks to ENEA support. For Objective 4 (to characterize the study sites for physicochemical, trophic and microbiological aspects), measurements of temperature, salinity, fluorescence, pH were conducted by multiparameter probe; nutrient content, abundance of fecal contamination indicator bacteria, total prokaryotic abundance, its viable fraction and heterotrophic culturable fraction were estimated.

The biofilm-producing capacity of bacteria isolated from water,examined by UO 2- UniME, appears greater at RB station impacted by sewage treatment plant effluent than at PTS, and increases at 12 months compared to 9 months. Strains isolated from -20m have greater ability to form biofilms than those from -5m. Antibiotic susceptibility profiles and resistance/sensitivity rates of water-isolated bacteria (UO2) show % of resistance between 62.73 (RB-20m) and 72.16% (TB-20m) of the total, that of susceptible bacteria between 26.09% and 17.05% at the same sites. The finding of high antibiotic resistance in strains isolated from areas close to the Zucchelli base confirms the relationship with human presence. However, antibiotic-resistant bacteria were also isolated from TB and AG, suggesting the presence of a natural "resistome" even in sites not exposed to contamination.

Phytoplanktonic communities showed values ranging from 85 (2nd campaign, AG site) to 1,209 x 10^3 cells/L (1st campaign, RB site). During the first year, the PTS and RB stations about two months apart (Nov 2017 and Jan2018) show the absolute highest phytoplanktonic concentrations. During the first campaign microalgae species other than Diatoms and Dinoagellate, classified as "other" represent the most abundant component(5.9-85.8%), while Dinoagellate(0.1-24.4%) are important only in RB5. Diatoms (10.8-94.0%) reach very high % in PTS and RB sites (second sampling). In the second campaign, the community composition is very different, with Diatoms dominating(66.7-99.4%) compared to Dinoagellate(0.2-31.0%) and "other" microalgae(0.4-15.5%) and with higher percentages at the RB site.  Enzymatic measurements (leucin aminopeptidase-LAP, beta glucosidase BGLU and alkaline phosphatase-AP) show high proteolytic rates at sites affected by sewage treatment plant effluent (RB) and low salinity water (AG), confirming how organic polymers stimulate microbial degradation (Fig 23). The capacity of the microbial community to utilize organic substrates estimated on Biolog plates is highest at the RB site at -20m, while the PTS site at -20m has low metabolic levels. Carbohydrates and amino acids are the most metabolized compounds at -5m, while at -20m Carbon- Phosphorus compounds and amino acids. Metabolic activity toward carboxylic acids and complex carbon compounds is observed at all sites; at the -20m PTS site, amines are not metabolized.

Biofilm matrix

The abundance of prokaryotes in the biofilm shows, as in water, higher concentrations in the RB site than in PTS, peaking after 2.5 months; higher percentages of dead cells are also observed, but in the PTS site the viable fraction increases from 9 to 12 months on PVC and PE. Viable cells are also present in TB and AG, especially on PE. In comparison with PE, PVC substrate allows the development of a higher abundance of heterotrophic bacteria (Fig. 26), especially in RB; 108 strains (90 from PVC, 18 from PE) were isolated in the 1.a campaign, 224 strains (189 from PVC, 35 from PE) were isolated in the 2.a, Regarding the ability of isolates to produce biofilms, PE substrate seems to favor colonization more than PVC. Antibiograms and antibiotic resistance/sensitivity rates of bacteria isolated from biofilms show a higher average incidence of antibiotic resistance (strong+intermediate, R+ I) than isolates from water. Strains isolated from biofilms on PVC, in RB at -5m show resistance rates from 36.7 to 56.8% increasing from 9 to 12 months. In Tethys Bay higher average resistance percentages are observed in the TB-20m control compared to the AG-20m impact exposed to melting. In summary, the nature of the substrate appears to influence the incidence of resistance, and it appears that: 1) The microbial community of biofilm on PVC shows in the RB-5m impact station discrete rates of antibiotic resistance already after 9 months, increasing after 12 months; 2) Bacteria isolated from biofilm grown for 12 months on PVC in the AG site show higher rates of antibiotic resistance than in the TB control. The PE substrate compared to PVC seems to favor the spread of antibiotic resistance, suggesting a higher adsorption capacity of contaminants (antibiotics) by this material. At greater depth (RB -20 m vs -5 m) the % of antibiotic-resistant bacteria is found to be increased and on PVC after 12 months is greater than strains isolated at 9 months.

The composition of the microbial community on PE and PVC (UO1, CNR-IRSA RM, Dr. Fazi) was analyzed by CARD-FISH under an epifluorescence microscope after incubating portions of the biofilm with specific oligonucleotide probes and calculating the percentages of the various microbial groups compared to the total abundance obtained by DAPI staining. The CARD-FISH technique allowed characterization of the microbial community attached to PVC and PE substrates. The Bacteria domain accounts for 82% to 92% of the total, with abundances on the order of 10^6 cells/cm2 and with higher abundances on PE. High diversity is found in the biofilm community with dominance of Proteobacteria and especially Alpha-Proteobacteria (27-37%) over Delta-, Gamma- and Beta-Proteobacteria (11, 12% and 8% respectively). Bacteroidetes make up max 17% of the total, while Actinobacteria and Firmicutes make up 12.8% and 2% respectively. The community in the waters appears less diverse and dominated by Bacteroidetes. Differences between sites are mainly observed for Gamma-Proteobacteria, abundant in RB, a station impacted by effluent from the Zucchelli base. The 3D structure of the biofilm was examined by laser scanning confocal microscope (CLSM) revealing already after two months the formation of a complex microbial biofilm, about 15 mm thick and with variegated surface. For microbial diversity analysis by high-throughput 16S rRNA gene sequencing, DNA was extracted from both water (500 ml) and biofilm (0.02 g) through PowerSoil DNA Isolation Kit. Shannon H' diversity index ranged from 4.8 to 5.9 for biofilm, 2.8 to 2.9 for water, with higher diversity in number of OTUs for biofilm vs. water (at RB site 4573 OTUs in water vs. 39224 in biofilm). The water and biofilm communities at the two sites RB and PTS are significantly different. In fact, in water, Bacteroidetes are more abundant (70% of OTUs); Alpha-Proteobacteria constitute about 60% of OTUs in biofilms, in water only 25%. Bacteroidetes constitute 9.6% of OTUs in RB and 16.1% in PTS. The genus Polaribacter of Bacteroidetes along with Psychroserpens and Ulvibacter, appears to be representative phylum in water, while Lewinella is the representative one in biofilm. The Alpha-Proteobacteria genus Sulfitobacter is found in water and biofilm, while Loktanella, Planktomarina and Amylobacter are found more in water.

Qualitative-quantitative analysis of the microalgal biofilm community by Utermohl method shows higher values on PE panels (range: 5.7-6915.4 x 10^6 cells/cm2) than on PVC (102.0-1013.4 x 10^6 cells/cm2). Moreover, in the second year, the highest algal concentrations are reached at station RB on PVC and PE. As for the qualitative composition, the microalgal communities are found to consist mainly of Diatoms(66.7-99.4%) while Dinoagellate(0.2-31.0%) and "other" microalgae(0.4-16.6.0%) rarely reach significant %. Compared with the water matrix, the taxonomic composition of the biofilm shows less diversity, with few species adapted to colonize the substrate.

As for microbial metabolism, higher rates of enzyme activity are found on biofilm after 2.5 months on PE than on PVC; after 9 and 12 months activity reaches higher levels on PVC and in the Road Bay area. The microbial community on PVC shows active ability to metabolize organic substrates in the Road Bay area at the RB site at 5 m with increases from 9 to 12 months, while in the Tethys Bay area metabolic activity levels are highest at the TB site at -20 m. Peak metabolic activity is observed as early as after 48 h of incubation. Except for the RB-20 m and PTS-5m sites, on PE the community shows less metabolic activity than on PVC. Complex carbon sources and C-P compounds are preferentially metabolized in TB, while more diverse metabolism is observed in AG.

Regarding Objective 2, the abundance and diversity of macrobenthos developed on PVC and polyethylene panels were studied and possible correlations with environmental physicochemical factors were evaluated. Artificial structures with panels fixed during the XXXIII campaign (2017-2018) were recovered in the following campaign (2018-2019) after 12 months (Long-term experiment). Only in the Road Bay area, 2 additional structures (frames) were fixed at -5 m, which were replaced after 2.5 months (February 2018, Short-term experiment), in order to obtain indications of summer and winter colonization processes separately. All panels, regardless of immersion time, were photographed as soon as they were recovered; after the fauna visible to the naked eye were taken and stored in ethanol at 96° at -18°C, they were dissected for all OTUs. The remaining parts with the remaining biofilm were air-dried and stored at -18°C. In Italy, the panels were allowed to air dry, then photographic documentation was made to assess the % coverage of each panel and the color of the biofilm, indicative of the type of coverage e.g., whether sediment or algal growth. For coverage analysis, photographs were processed using Imagej program, first calculating the area covered by biofilm and then any 3D biofilm structures ("overlapping"). In RB in the panels at -5 m for 12 months, exposed to sedimentation from the column, the biofilm layer is about 3 times that of the other panels, with overlapping greater than 100%. Also for color analysis, the ImageJ program allowed reconstruction of the spectrum in RGB; the parts of the panel not covered by biofilm were discarded and the obtained histograms grouped to assess any differences by site, depth, and immersion time. To identify the fouling macrofauna, a detailed photographic analysis was carried out by light microscope first, followed by direct counting. Organisms preserved in ethanol were added to the obtained macrofauna list, reporting the density of individuals at 1 cm2. Electron microscopy preparations (stubs) were set up for some organisms, then gilded and photographed by SEM, allowing some specimens to be recognized at the species level. Data for the major taxonomic groups (fouling communities) were examined by site and depth. At sites RB and TB, for the same dive time, more biofilm developed at -5 m than at -20 m. In the PTS site, on the other hand, the cover appears poor, except at -5 m after only 2.5 months. The failure to recover panels in site AG {5 m (gone missing) unfortunately does not allow for generalized conclusions. The large overburden in the panels at site RB -5 m after 12 months could also be due to the forced exposure of only one side of the panels to the overlying water column, due to overturning of the panels caused by sea currents. The trend of the colonization process at the different sites shows that at the different sites and at the different depths the biofilm weight follows a similar trend as the cover, except in PTS, where cover and biofilm weight show opposite trends, indicating that the final biofilm weight is significantly affected by the percentage of organisms weighing more than diatom frustules. The values of the number of macrozoobenthic individuals, in fact, reflect the trend of biofilm weight values.

Biofilm color analyses reveal no specific differences except for a few panels:

-Pvc panels from site PTS-5m after 2.5 months are characterized by a bright green overlay, possibly due to a microalgae with `mathed' overlay.

-Panels from site AG-20m (submerged for 12 months under the Amorphous Glacier saracack in Tethys Bay) whose grayish color was due to predominant accumulation of fine sediment perhaps from the glacier, as opposed to diatoms.

-Panels submerged for 12 months at RB, both at -5 and -20 m, are characterized by a pronounced yellowish color due to both the diatoms present and the large amount of fine sediment, markedly different from that of AG. In addition to the fact that the bottom of Road Bay has a finer sediment than the other study sites (all reef promontories), fine sediment also comes from the sewage treatment plant discharge.

Road Bay is a polluted bay, where poor circulation and the presence of organic discharges could explain the proliferation of diatoms and nematodes on the panels. harpacticoids are grazers that depend on algal cover for their sustenance, except at the PTS -5 m site after 2.5 months where reduced grazing pressure would explain the presence of intact algal biofilm. This could be due to failure to reach the climax (=equilibrium in an ecological succession) where grazers development is observed and secondary production increases. In contrast, the phase shift between producers and consumers is not distinguishable after 9 and 12 months, being masked by annual variability. Ostracods show with nematodes higher abundances in RB after 9 and 12 months, suggesting marked eutrophication due to sewage discharge; in reality this is an artifact attributable to the detachment and fall of panels to the bottom. In this case, instead of actual colonization of the panels, there was a deposition of detrital material from the water column, which caused an increase in biofilm and sediment adhered to it, attracting consumers. This hypothesis appears to be supported by the presence of these groups even at sites with clean water (AG and TB). An interesting result is the almost complete absence of ostracods and nematodes in the PTS site, which is strongly poor in primary and secondary productivity. No bivalves are present in the PTS site after 2.5 months, suggesting how the establishment of their larval stages (veliger) occurs in late summer. The benthic community present on the panels that remained submerged during the winter appears poor in Sabellids and Hydrozoa. Gastropods are vagile and seem to have been attracted from late summer onward by the presence of food, as indicated by their presence in PTS and RB at 9 and 12 months, but especially their absence after 2.5 months; they also seem to prefer clean sites, given their low abundance in RB -5m at 12 months. Low productivity is observed at the PTS site, except for Sabellids, Hydrozoa and gastropods, the latter being well represented here. At RB -5 m filterfeeders (Sabellids, Brachiopods, Hydrozoa,Sponges) are missing or low in abundance, perhaps suffering from excessive sedimentation from the column when the panels had upward-facing faces. The remaining sites (RB-20, TB-5, TB-20, and AG-20), with more transparent waters probably favor phytoplanktonic organisms such as Brachiopods and Bivalves. A separate observation concerns the recruitment of Adamussium colbecki, which occurred only at the -20 m sites, and that of the Brachiopod Craniata sp. and the sponge Stylocordyla chupachups, which will allow a new morphological description of benthic species.

Regarding Objective 3, to explore the ability of microbial isolates to produce biofilms and secondary metabolites and to secrete hydrolytic enzymes, bacterial strains were isolated from the biofilm developed on PVC panels submerged in Nov 2017 at -5 m at RB and PTS. In the first campaign, 82 colonies were isolated from the biofilms collected after 2.5 months, 17 with morphology of filamentous actinomycetes and 8 of filamentous fungi, with the remaining classified as bacteria. In the second campaign (Nov-Dec 2018) biofilm samples were seeded under the best isolation conditions of the first campaign (TQ, ISP4 and Marine agar, incubation at RT and 4°C, no treatment) and 64 colonies were isolated (Fig. 55). A total of 146 strains were isolated, including 28 filamentous actinomycetes and 10 filamentous fungi. All isolates grew at both RT and 4°C and 28.1% as obligate halophiles. To identify microorganisms with the greatest potential as producers of hydrolytic enzymes of potential biotechnological and/or industrial interest, assays were conducted for protease, lipase, amylase, cellulase and chitinase activities. Of the 146 isolates, 92 (63%) of which 22 actinomycetes (78.6% of actinomycetes) and 9 fungi (90% of fungi) showed at least one activity. Forty bacteria isolated from RB -20 m in Gen 2018 from UO2 were also subjected to the same screening, of which 39 (97.5%) were active for at least one enzyme hydrolysis assay: 33 positive for protease activ., 38 lipase, 14 amylase, none for chitinase or cellulase. This study represents a wealth of useful information for characterizing some potential extremophilic microorganisms and enzymes resistant to high salt concentrations and with good activities at low temperatures. A screening to identify oxidative laccase and/or peroxidase enzyme activities useful for degrading recalcitrant compounds, was conducted on MAM medium in the presence of ABTS or Azure B. Of the 40 marine isolates, 17 tested positive on ABTS and 11 on Azure B. Of the 146 isolates from biofilms, 8 were positive on ABTS (5 fungi and 1 actinomycete), while 9 isolates (5 fungi and 2 actinomycetes) had activity on Azure B. The positive bacterial isolates were then cultured in LB + 2% NaCl liquid medium. Five marine bacteria showed activity on ABTS, catechol, 2,6-DMP and Azure B. Isolates from biofilms, active in initial screening and filamentous growth (known producers of oxidative activities), cultured in SSC only supernatants of 4 fungi showed activity on at least one substrate. For the production of antibacterial and antifungal metabolites, only actinomycetes and filamentous fungi, the main producers of secondary metabolites, were tested. Seven actinomycetes with activity on B. subtilis were then grown in V6 and SSPM; all grew optimally on SSPM and showed activity on S. aureus. Broculture and supernatant of strain 255S also generated halos of inhibition toward E. coli and P. tomato. No isolates showed antifungal activity. The antibiotic resistance profiles of the isolates provided not only useful information on the spread of antibiotic resistance but also for the search for antibacterial molecules. To assess possible antibiotic resistance, the 136 isolates (except fungi) were plated in the presence of antibacterial molecules: 90 isolates (66.2%) were found to be resistant to at least one of the antibiotics used. For all classes of antibiotics, more resistant microorganisms were observed at the RB site, demonstrating the inuence of anthropogenic impact on the spread of bacterial resistance.

Articles in scientific journals

1)Caruso G., Azzaro M., Dell'Acqua O., Lo Giudice A., Fazi S., Caroppo C., Azzaro F., La Ferla R., Maimone G., Laganà P., Marinelli F., Berini F., Marcone G.L., Pichon G.,Chiantore M. (2018) THE ANT-BIOFILM PROJECT (PNRA): Biological colonization of Antarctic coastal sites and biotechnological prospecting. Biologia Marina Mediterranea, 25 (1): 267-268 Finanziato per il 100% da PNRA16 00105 (ANT-Biofilm)

2)Caruso G., Azzaro M., Dell'Acqua O., Lo Giudice A., Fazi S., Caroppo C., Azzaro F., La Ferla R., Maimone G., Laganà P., Marinelli F., Berini F., Binda E., Raffa F.,Chiantore M. (2019) Microbial response to anthropic and natural forcings in two coastal Antarctic sites (Ross Sea). Biologia Marina Mediterranea, 26 (1): 379-380 Finanziato per il 100% da PNRA16 00105 (ANT-Biofilm)

3)Laganà P., Caruso G.,  Corsi I., Bergami E., Venuti V., Majolino D., La Ferla R., Azzaro M., Cappello S. (2019) Do plastics serve as a possible vector for the spread of antibiotic resistance? First insights from bacteria associated to a polystyrene piece from King George Island (Antarctica). International Journal of Hygiene and Environmental Health, 222: 89-100 Finanziato per l'80% da PNRA14_00090 (PLANET); per il 10% da PNRA16_00105 (ANT-BIOFILM) e per il 10% da PNRA 16_00075 (nanoPANTA)https://doi.org/10.1016/j.ijheh.2018.08.009.

4)Lo Giudice A., Caruso G., RIzzo C., Papale M., Azzaro M. (2019) Bacterial communities versus anthropogenic disturbances in the Antarctic coastal marine environment. Environmental Sustainability, 2: 297-310. https://doi.org/10.1007/s42398-019-00064-2 Finanziato da PNRA (PNRA16 00020, progetto P3 per il 50%) e da PNRA16 00105 (progetto ANT-Biofilm per il 50%)

5)Caruso G (2019) Microplastics as vectors of contaminants. Marine Pollution Bulletin, 146: 921-924. https://doi.org/10.1016/j.marpolbul.2019.07.052

Finanziato per il 100% da PNRA16 00105 (ANT-Biofilm)

6)Caruso G (2020) Microbial Colonization in Marine Environments: Overview of Current Knowledge and Emerging Research Topics. Journal of Marine Science and Engineering (MDPI), 8, 78; doi:10.3390/jmse8020078

Finanziato per il 100% da PNRA16 00105 (ANT-Biofilm project).

7) Cappello S., Caruso G., Bergami E., Macrì A., Venuti V., Majolino D., Corsi I. (2021). New insights into the structure and function of the prokaryotic communities colonizing plastic debris collected in King George Island (Antarctica): preliminary observations from two plastic fragments. Journal of Hazardous Materials, 4141; 125586. DOI 10.1016/j.jhazmat.2021.125586  Finanziato per il 70% da PNRA \Plastic in Antarctic Environment" (PLANET, PNRA 14 00090) e per il 30% da PNRA16 00105 (ANT-Biofilm)

8)Caruso G., Papale M., Azzaro M., Rizzo C., Laganà P., Lo Giudice A. Antarctic sponge homogenates as a source of enzymes and antibacterial substances. Polar Biology, submitted.

9)Fazi S., Dell'Acqua O., Azzaro M., Venuti V., Fazio E., Fazio B., Papale M., Lo Giudice A., Caruso G. Structural and molecular composition of biofilm microbial communities colonizing plastic substrates in Terra Nova Bay (Antarctica), in preparation