Antimicrobial Activity of Heterotrophic Bacterial Communities from the Marine Sponge Erylus discophorus (Astrophorida, Geodiidae)

Heterotrophic bacteria associated with two specimens of the marine sponge Erylus discophorus were screened for their capacity to produce bioactive compounds against a panel of human pathogens (Staphylococcus aureus wild type and methicillin-resistant S. aureus (MRSA), Bacillus subtilis, Pseudomonas aeruginosa, Acinetobacter baumanii, Candida albicans and Aspergillus fumigatus), fish pathogen (Aliivibrio fischeri) and environmentally relevant bacteria (Vibrio harveyi). The sponges were collected in Berlengas Islands, Portugal. Of the 212 isolated heterotrophic bacteria belonging to Alpha- and Gammaproteobacteria, Actinobacteria and Firmicutes, 31% produced antimicrobial metabolites. Bioactivity was found against both Gram positive and Gram negative and clinically and environmentally relevant target microorganisms. Bioactivity was found mainly against B. subtilis and some bioactivity against S. aureus MRSA, V. harveyi and A. fisheri. No antifungal activity was detected. The three most bioactive genera were Pseudovibrio (47.0%), Vibrio (22.7%) and Bacillus (7.6%). Other less bioactive genera were Labrenzia, Acinetobacter, Microbulbifer, Pseudomonas, Gordonia, Microbacterium, Micrococcus and Mycobacterium, Paenibacillus and Staphylococcus. The search of polyketide I synthases (PKS-I) and nonribosomal peptide synthetases (NRPSs) genes in 59 of the bioactive bacteria suggested the presence of PKS-I in 12 strains, NRPS in 3 strains and both genes in 3 strains. Our results show the potential of the bacterial community associated with Erylus discophorus sponges as producers of bioactive compounds.


Introduction
Due to their physico-chemical properties, which rarely exceed the biological tolerance limits, the oceans provide a safe environment for most living organisms [1]. However, the multiple beings that live in the marine environment had to develop survival strategies against other organisms with whom they have to compete for space and food. Sharing a common environment over a long evolutionary period also allowed the establishment of well-balanced associations between many of these organisms. A good example of these associations are the sponges that host a significant microbiome which can reach up to 40-60% of the total sponge biomass and densities of 10 8 to 10 10 bacteria per gram of sponge wet weight. These values can exceed seawater concentrations by two to three orders of magnitude [2][3][4][5].
Natural bioactive compounds have been used since the beginning of traditional medicine [6]. They are present in all kinds of life forms and are produced by the secondary metabolism of organisms. Secondary metabolites include terpenoids, alkaloids, polyketides, peptides, shikimic acid derivates, sugars, steroids and a large mixture of biogenesis metabolites [7]. Sponge symbionts are fundamental in host defence against predators due to the production of biologically active secondary metabolites. These natural products can show antibacterial, antifungal, antiviral, antiprotozoal, anthelmintic, anti-inflammatory, antitumor, immunosuppressive, neurosuppressive properties and can also possess activities for the treatment of cardiac, respiratory and gastrointestinal diseases [8]. Sponges are the ''gold mine'' organisms for natural product isolation (over 30% of the products) in the marine environment [9]. The search for new drugs, especially antibiotics, is important due to the increase of bacterial resistance to existing antibiotics.
The true origin of many of these bioactive molecules is uncertain. The production of secondary metabolites could be due to the cooperation between sponges and symbionts, only to the symbionts or to the sponges [10;11]. A microbial origin got support from the occurrence of structurally similar substances in unrelated sponges (see Laport et al. [8]). The fact that these substances may be produced by microbes could allow their sustainable and unlimited production in vitro. This is hardly the case with sponges. As the sponge bioactivity may in fact be due to their microbiome, these organisms became the subject of many works.
Secondary metabolites possess complex structures and involve unusual biochemistry. Two families of enzymes, the polyketide synthases (PKS) and the nonribosomal peptide synthetases (NRPS), are of particular importance in the production of various secondary metabolites many of which are important drugs [12]. Both PKSs and NRPSs can be conceptualized as enzymatic ''assembly lines'' composed of functional modules [13].
The discovery of new biosynthetic pathways encoding genes of secondary metabolites opens the possibility of heterologous production and the genetic manipulation of the gene cluster to obtain new natural products [14]. Metagenomic analysis of the prevalence and presence of PKS and NRPS genes are being studied to improve the search of new bioactive compounds in sponges (e.g. Schirmer et al. [15]).
In this study we aimed to assess the bioactive potential of bacteria isolated from the marine sponge E. discophorus collected in Berlengas Islands against a panel of pathogenic and environmental microorganisms. Antimicrobial bioactivity was detected in 31% of the 212 isolated heterotrophic bacteria from two specimens of E. discophorus and the presence of PKS-I and NRPSs genes was detected in several isolates showing the biotechnological potential of these bacteria.

Biological material
The 212 bacteria (here designated test bacteria) under study were isolated from two specimens of the sponge E. discophorus (named Berg01 and Berg02) that were sampled nearby, at 15 m by scuba diving in the Berlengas Islands a protected area of UNESCO located off the W coast of Portugal (N39u 289 470; W9u 329 780). The authors thank Reserva Natural das Berlengas (Instituto da Conservação da Natureza e da Biodiversidade -ICNB) for the sponges samples harvesting permission. Voucher samples of the sponges were preserved in 90% ethanol for taxonomic identification and deposited in the Biology Department's zoological collection of the University of the Azores (DBUA.Por). Specimens were identified from the analysis of general external and internal morphological characters, i.e. shape, type, size and arrangement of skeletal structures (spicules) following the Systema Porifera classification system [33].
The bacterial isolation was performed inoculating serial dilutions (10 21 to 10 26 ) of the homogenized from the sponges in heterotrophic media. After isolation in pure culture, bacteria were cultivated in Marine Broth (Becton Dickinson, MB) in the dark at 26uC. Their taxonomic identification was based on the analysis of the 16S rRNA gene by direct colony PCR or with template DNA extracted by the boiling method. A loop full of bacterial culture was resuspended in 200 ml of distilled deionized H 2 O and subjected for 10 min to 100uC, cooled on ice and the extracted DNA amplified with the universal primers 27f and 1492r [34] in 50 ml of PCR mixture (16 PCR buffer; 1.5 mM MgCl2; 1 unit of GoTaq Flexi DNA Polymerase; 200 mM of each deoxynucleoside triphosphate (dNTPs); 2 mM of each primer). The PCR program was performed in a MyCycler TM Thermo Cycler (Bio-Rad) and consisted in an initial denaturing step of 5 min at 95uC; 30 cycles of 1 min at 94uC; 1 min at 52uC; and 90 s at 72uC; and a final extension of 5 min at 72uC. PCR products were visualized after electrophoresis in a 1.2% agarose gel in 1X TBE buffer. The PCRs amplicons were sequenced at MACROGEN (Korea). Sequences were assembled with Vector NTI 10.1 and classified in the Ribosomal Database Project [35].

Screening assays for antibacterial and antifungal activity
The search for bioactivity was performed with the 212 bacteria isolated from E. discophorus. The isolates were initially fermented in 96-deep well Duetz plates [36] with 800 ml of MB media in each well for 3 days at 28uC and 220 r.p.m. After incubation, 400 ml of each culture were transferred to new 96-deep wells plates containing 400 ml of a solution of Sea Salts (4%) and Glycerol (40%) and kept at 228uC. These plates were designated as the Master Plates (MP). The MPs were then used in a replication procedure previously described [36] to generate fresh inocula (MB medium, 2 days, 28uC, 220 rpm, 70% humidity). The inocula thus generated were employed to seed all subsequent micro-fermentations carried out in both Janus and Duetz plates.
2.1 Janus plates assays. This double-faced plated assay optimized by Fundación MEDINA [37] is based in exclusively designed plates by Nunc with two sides, which allow culture growth on the opposite layers in solid media. In order to maximize the number of potential active secondary metabolites produced, 5 different media were chosen to carry out the miniaturized fermentations: Marine Broth (2216 Difco), Medium F (0.015 g K 2 HPO4; 0.2 g CaCl 2 ; 0.75 g KCL; 23.4 g NaCl; 7 g MgSO4; 1 g Mannitol; 1 g Yeast extract; 1 g Peptone; 16 g Agar; 1 ml Hutner's basal salts [38] and 999 ml ddH 2 O), R2A (218263 -Difco) and saline (supplemented with 3% of SeaSalts of Sigma) R2A and Starch Agar (Difco).
On the top layer of the Janus plates the test bacteria were inoculated employing a replication procedure previously described [36] and incubated for 3 days at 28uC and 70% of humidity. Subsequently, the inoculated media with the target microorganisms were added on the opposite side thus forming the assay layer. The double-faced Janus plates were incubated at 37uC for 20-24 h. A search of inhibition zones indicative of antibacterial or antifungal activity was then carried out [37]. This in vitro screening assay was performed against a panel of Gram positive bacteria (Bacillus subtilis (ATCC6633) and Staphylococcus aureus MRSA), Gram negative bacteria (Acinetobacter baumannii) and the yeast Candida albicans. All target organisms were pre-inoculated and grown at 28uC, 220 r.p.m. for 12 h and 70% of humidity. Preinoculum and inoculum medium used to grow A. baumannii and B. subtilis was Luria Broth (LB) (Miller's, Invitrogen). For C. albicans the pre-inoculum medium was Sabouraud Dextrose Broth (SDB) and the inoculum medium was YNBD [37]. The pre-inoculum and inoculum medium prepared for the overnight growth of S. aureus MRSA was Brain Heart Infusion (BHI). Inocula absorbance for all target microorganisms was adjusted to a final 600 nm optical density (OD) of 0.4 before being placed in the Janus plates. Non-inoculated medium was used as negative control.
Tests of growth interference of the target microorganisms with saline media were performed in advance to rule out possible interferences.
2.2 Microfermentations in 96-deep well plates. MPs were used to carry out microfermentations in 96-deep well plates (here designated by Duetz system assay) following the approach described by Duetz et al. [36]. For this assay, besides the 5 media already specified for the Janus plates, half saline concentration of the media were also tested. The inoculated Duetz plates were incubated (1 mL) for 5 days at 28uC, 220 r.p.m. and 70% of humidity. Bacterial broth were then subjected to an organic extraction with 800 ml acetone and 40 ml DMSO per well. The plates were incubated for 1 h at 220 r.p.m. and then transferred to a vacuum centrifuge (GeneVac HT-24) in order to reduce the final volume to 400 ml (2 Whole Broth Equivalent). The supernatants of the extracts solutions were transferred to 96-deep well plates. The organic extracts were then assayed against the Gram-negative Pseudomonas aeruginosa PAO1, Acinetobacter baumannii, Vibrio harveyi (CECT 525) and Aliivibrio fischeri (CECT 524); and the Grampositive Bacillus subtilis (ATCC6633), Staphylococcus aureus wild type Smith strain and Staphylococcus aureus MRSA. Antifungal assays were also prepared against the yeast Candida albicans and Aspergillus fumigatus (ATCC46645) and (DakuB KU80 ). Unless specified negative controls were always uncultivated culture medium.
For the screening assays against Acinetobacter baumannii, S. aureus (Smith), Pseudomas aeruginosa PAO1 and Candida albicans, overnight cultures grown in liquid Luria Broth (Miller's media) at 37uC and 220 r.p.m. were measured at an absorbance of 612 nm for A. baumannii and S. aureus (Smith) and 600 nm for P. aeruginosa. The inocula in LB media were adjusted to an OD of 5610 5 for A. baumannii and to a cell concentration of 2.5610 5 CFU/ml for S. aureus (Smith) and 5610 5 CFU/ml for P. aeruginosa. A suspension of C. albicans in medium RPMI (40 ml 1 M HEPES; 36 ml 50% glucose; 6.7 g Yeast Nitrogen Broth without amino acids; RPMI a bottle; adjust pH to 7.1 and 0.22 mm sterilized) was adjusted to 0.250 at 660 nm. This suspension was then diluted 1/10. The plates were incubated at 37uC for 20 h under humid conditions. The absorbances were measured at 612 nm in a Tecan ULTRAEVOLUCION before and after incubation. To confirm the results, 30 ml of a 0.02% resazurin stock solution was added to each well (100 ml) and incubated for 2 h. Changes in colour from blue (growth inhibition corresponding to detection of bioactivity) to pink (no growth inhibition corresponding to no detection of bioactivity) and fluorescence readings radiated from the resazurin were measured in a Perkin Elmer VICTOR2 multi-function fluorometer. All ODs and fluorescence measurements were analysed using the Genedata Screener software.
For the screening assays against Aspergillus fumigatus ATCC46645 and DakuB KU80 , cultures of both A. fumigatus were prepared in medium RPMI with 0.002% resazurin from a stock spore suspension in Tween saline buffer to a final concentration of 2.5610 4 conidia/ml as described by Monteiro et al. [39]. The assays were carried out as previously described, using as positive controls 0.5, 1.0, 2.0 and 4.0 mg.mL 21 amphotericin B. The plates were incubated for 30 h at 37uC and, then, the fluorescence was recorded in a Perkin Elmer VICTOR2 2TM Multi-function fluorometer.
For the screening assays against Staphylococcus aureus MRSA, an overnight culture of S. aureus MRSA in 10 ml of liquid Brain Heart Infusion (BHI) medium was incubated at 37uC and 220 r.p.m. The OD of the culture measured at 660 nm was adjusted to 0.2 and used to inoculate BHI agar medium (3 ml of the adjusted inoculum per 100 ml of medium) which was then distributed (30 ml) in OmnyTray plates. Disposable 96 pin trays were used to generate 96 wells in each of the BHI agar plates in which, subsequently, 10 ml of each extract in each well were dispensed. Alternatively extracts were distributed in BHI agar plates without wells. Ten ml of 0.5 mg.mL 21 kanamycin was used as positive control. Plates were then incubated at 37uC for 20 h and zones of inhibition (ZOI) were measured. Any extract producing a visibly discernible ZOI, regardless of zone quality, was considered to be positive.
For the screening assays against Bacillus subtilis (ATCC 6633), a culture of B. subtilis was prepared using 1 ml of spore suspension/ 1 L medium (23 g/L of Nutrient Agar and 2 g/L of yeast extract) that had been previously sonicated for 3 min. The assay was performed in a similar way to the one used against S. aureus MRSA and the positive control was 150 mg/ml tunicamycin.
Screening assays against Vibrio harveyi CECT 525 and Aliivibrio fischeri CECT 524 were carried out in a cell density optimized agar assay. Ten ml of Luminous Medium [40] were inoculated with a loop of the pure V. harveyi and A. fischeri, incubated at 25uC, 220 r.p.m. for 12 h and the optical density adjusted to 0.3 for V. harveyi and 0.4 for A. fischeri at 600 nm and, subsequently, both diluted by 1/10. The assay was performed in a similar way to the one used against S. aureus MRSA and the positive control consisted on 0.256 mg/ml chloramphenicol. The cultures were incubated at 25uC for 24 h. Inhibition was detected with the presence of nonphosphorescent halos in a dark-room.

Search of PKS and NRPS genes
The molecular analysis of the genes PKS-I and NRPS involved in the production of secondary metabolites was investigated in the 59 of the bioactive bacteria. Genomic DNA was extracted with an E.Z.N.A. bacterial KIT from OMEGA. Specific degenerated primers MDPQQRf and HGTGTr [41] and DKf and MTr [42] were used for PCR amplification of PKS-I and NRPS genes, respectively. A total of 25 ml of PCR mixture (16PCR buffer with 1.7 mM MgCl2; 0.8 unit of Go Taq DNA Polymerase; 0.2 mM of each dNTPs; 0.1 mM of each primer and 5 ml DNA template) was used. The PCR program for the genes PKS-I and NRPS was the same and consisted of an initial denaturing step of 5 min at 95uC; 11 cycles of 1 min at 95uC; 30 s at 60uC and 1 min at 72uC, with the annealing temperature reduced by 2uC per cycle, followed by 30 cycles of 95uC for 1 min, 40uC for 30 s and 72uC for 1 min with a final extension of 10 min at 72uC. The PCR programs were performed in an MyCycler TM Thermo Cycler (Bio-Rad) and the PCR products were visualized in VWR GenoPlex after electrophoresis in a 1.2% agarose gel in 1X TAE buffer.

Results and Discussion
The oceans, an almost endless source of microbial diversity, are the habitat of organisms such as sponges that harbour a large microbial diversity with important biosynthetic potential due to their secondary metabolites profiles [43]. The analysis of sponge symbionts in pure cultures is an advantage for the performance of bioactive screening assays [44] and is the most direct method for the large-scale production of bioactive compounds [45]. The two specimens of Erylus discophorus collected in Berlengas (Berg01 and Berg02) allowed the isolation of a large number of heterotrophic bacteria (212 isolates) of which 31% (66 isolates) showed bioactivity. Of the screened bacteria for bioactivity and based on the 16S rDNA gene analysis, 57% (n = 120) were Alphaproteobacteria, 21% (n = 45) Gammaproteobacteria, 16% (n = 34) Actinobacteria and 6% (n = 13) Firmicutes.
Bioactivity was observed against one or more of the target microorganisms tested. The majority was active against Bacillus subtilis (87.9%) and at a lower percentage against Staphylococcus aureus MRSA (10.6%), Aliivibrio fischeri (9.1%) and Vibrio harveyi (6.1%). Bacteria with bioactivity against both B. subtilis and S. aureus MRSA represented 9.1% and against both A. fischeri and V. harveyi accounted for 4.6%. No bioactivity was observed against P. aeroginosa, A. baumannii, S. aureus wild type, C. albicans and A. fumigatus.
The taxonomic affiliation of all bioactive isolates is provided in Table 1. Table 2 correlates the taxonomic position of the isolates genera and their relative bioactivity percentage. It also specifies the number and relative percentage of bioactive isolates obtained with the Janus and Duetz systems. Thirty two isolates of Alphaproteobacteria (48.5%) belonging to the genera Pseudobrivio and Labrenzia were bioactive against B. subtilis, S. aureus MRSA and A. fischeri. Eighteen isolates of Gammaproteobacteria (27.3%) belonging to the genera Vibrio, Acinetobacter, Microbulbifer and Pseudomonas were active against B. subtilis, V. harveyi and A. fischeri. Eight isolates of Actinobacteria (12.1%) belonging to the genera Gordonia, Microbacterium, Micrococcus and Mycobacterium were active against B. subtilis and S. aureus MRSA. Eight isolates of Firmicutes (12.1%) belonging to the genera Bacillus, Paenibacillus, Sporosarcina and Staphylococcus were active against B. subtilis, V. harveyi and A. fischeri. Pseudovibrio (47.0%), Vibrio (22.7%) and Bacillus (7.6%) are the three most bioactive genera of all the bioactive isolates. No activity was observed in isolates affiliated to Ruegeria, Rhodobacter, Erythrobacter, Martelella, Nautella, Photobacterium, Thalassomonas, Rhodococcus and Dietzia.
The group with the higher number of isolates demonstrating bioactivity was the Alphaproteobacteria followed by the Gammaproteobacteria, Actinobacteria and Firmicutes. However, if the analysis is made based on the number of bioactive isolates relative to the total number of bacteria in each group, Firmicutes are the most bioactive (61.5%) followed by Gammaproteobacteria (40%), Alphaproteobacteria (26.7%) and Actinobacteria (23.5%). Bioactivity results obtained with marine bacteria and sponge associated bacteria are somehow different. Regarding marine bacteria in general, most of the new marine bacterial compounds from 1997 to 2008 were originated from Actinobacteria (40%), Cyanobacteria (33%), Proteobacteria (12%) and Firmicutes and Bacteroidetes (5% each) [46]. The distribution of bioactive compounds produced by sponge associated bacteria is Actinobacteria (46.66%), Proteobacteria (23.33%), Firmicutes (11.66%), Cyanobacteria (8.33%), Verrucomicrobia (5%) and others (5%) [47]. The high number of bioactive Actinobacteria may be biased due to their extensive study in the production of antibiotic compounds since 50% of known microbial antibiotics are derived from these bacteria [48].
Several of the obtained bioactive genera are well known producers of metabolites with antimicrobial properties but others are less known. Furthermore, many of the examples referred to below are from sponge associated bacterial isolates.
A suite of antimicrobial compounds with spectra of different antimicrobial activity was observed in Pseudovibrio [11;49-52]. A sponge associated Alphaproteobacterium related to Pseudovibrio denitrificans, displayed a weak and unstable antimicrobial activity, which was easily lost during cultivation [53]. However, this bioactive bacterium was present in the sponges in high numbers. High antimicrobial activity was also observed in isolates from soft corals affiliated to the alphaproteobacterium Labrenzia [51].
Marine vibrios have been reported as a rich source of novel biologically active metabolites [10;51;54-57]. Bioactivity produced by Vibrio sp. and sponge extracts was observed against Bacillus [58]. A total of 93 secondary metabolites were isolated from Vibrionaceae [59]. These are surface-associated bacteria known to produce a broad range of antibacterial compounds which may have a relevant ecological role favouring their abundance in microbial communities [59].
Bacillus spp. from terrestrial origin are well known sources of antimicrobial compounds [69]. Similarly antibiotics/bioactive compounds from marine Bacillus spp. have also been reported [10;53;55-57;70;71]. Genome sequencing studies of the genus Bacillus have revealed its potential as a source of antibiotic-like compounds [72].
Micrococcus strains possessing antimicrobial activity were described by Bultel-Poncé et al. [61], Hentschel et al. [76] and Lo Giudice et al. [77]. Antitumor and antimicrobial bioactivity was observed by Microbacterium species [57;78;79]. In a recent review the actinomycete genus Gordonia was described as being capable of degrading xenobiotics, environmental pollutants, or otherwise slowly biodegradable natural polymers as well as to transform or synthesize possibly useful compounds [80]. However, to our knowledge, no antimicrobial activity has been associated with this genus, nor with the Mycobacterium genus which is well known for its infectivity.
The most efficient liquid medium for the production of bioactive compounds against B. subtilis, S. aureus MRSA, V. harveyi and A. fischeri was MB. In saline R2A bioactivity was observed against both S. aureus MRSA and V. harveyi, and in saline Starch and MF against A. fischeri. In solid media assays, higher numbers of inhibitions were observed in saline Starch followed by MF and saline R2A. No inhibitions were found in MA medium when used as culture medium in the Janus system. The non-saline R2A medium was the only one where no bioactive compounds were produced in both systems.
The screening of the 212 bacteria was performed using two different plate systems. The Duetz system is a miniaturized fermentation system that allows carrying out a high number of microfermentations with a lot less effort than the effort required to carry out the fermentations in tubes/flasks. Bacteria are fermented in liquid media, crude extracts are obtained and assayed for bioactivity against target organisms. However, in the double-faced diffusion assay Janus system, after the initial growth of the sponge isolates, the medium, containing diffused secondary metabolites is put in contact with the target organism to assess inhibitory responses. The Janus system allowed the assessment of a higher number of bioactive positive hits (84%) when compared to the Duetz system (16%). However, the Duetz results are more reliable than the ones obtained with the Janus plates due to the overlap of inhibition halos in the latter. Moreover, due to bacterial swarming and gliding, results in the Janus system were obtained after 3 days of incubation, whereas longer incubation times were employed in the liquid microfermentations. Bacterial swarming and gliding also interfered with the visualization of results in other antimicrobial studies [37;53].
It is well known that microbial secondary metabolite production is highly-dependent on the fermentation conditions [81]. Growth media and growth conditions are variables that are known to have effects on the production of bioactive compounds and can be different depending on the strains [82]. As the production of secondary metabolites is mainly observed in late exponential/ stationary phases, bacteria should produce more bioactive compounds after 5 days than 3. This hypothesis could be true for some bacteria and may explain some of the different bioactivity percentages obtained with the two methods. Furthermore, being     Janus plates assay a microbial culturing system, the production of secondary metabolites is possibly continued along the entire assay at least for some bacteria able to grow at 37uC (incubation temperature of the target microorganisms). Thus some bacteria might be able to produce bioactive compounds in a continuous way for longer while in direct contact with target microorganisms [37]. It is important to notice that assays performed with Duetz systems are cultivated in individual wells, oppositely to the Janus assays where the cultivation of the 95 bacteria was performed in the surface of agar without barriers. Therefore, there might be competition for nutrients, growth inhibition due to metabolites produced by the neighbours or even the synergic or antagonistic interaction of two or more bacteria in the production of secondary bioactive metabolites. Thus, this system is not as ''clean'' as the individual fermentations in Duetz. These aspects can also have some repercussion in the number of hits that were obtained with the two approaches The search for PKS-I and NRPS genes in 59 of the bioactive bacteria revealed that 30.5% (n = 18) of the bacteria amplified one or two of the genes for secondary metabolites. The presence of PKS-I was observed in 12 strains, NRPS in 3 strains and both PKS-I and NRPS in 3 strains (Table 3). The non-amplification of any of these genes in bioactive bacteria (69.5%) suggests that these strains should amplify with other specific primers for the PKS-I and NRPS genes or use other metabolic pathway for the production of secondary metabolites like PKS-II gene [83].
The majority of the bioactive bacteria that amplified PKS-I and NRPS genes belongs to the genus Pseudovibrio (n = 10). These genes were already reported in Pseudovibrio strains isolated from a Irciniidae sponge [84;85] Although a high number of the bioactive bacteria belongs to the genus Vibrio, only 3 bacteria amplified the PKS-I gene and none the NRPS gene. A recent review on the family Vibrionaceae suggests that only NRPS or hybrid PKS-NRPS genes were amplified [59]. Furthermore, the genus Gordonia (n = 2), Bacillus (n = 2) and Pseudomonas (n = 1) amplified the genes NRPS and PKS in lower numbers.
A wide range of bacterial groups were tested for the presence of the genes PKS and NRPS and they were found among other genera in Pseudomonas, Vibrio and Bacillus [93].
These results confirm the production of secondary active metabolites by some Erylus strains that are, thus, the most promising ones for future work.
The complex bacterial communities in marine sponges play a considerable ecological role in several aspects of the biology of these organisms, namely by the production of secondary metabolites fundamental for sponge protection against other organisms. These communities have thus a great biotechnological importance in the search for new and more effective pharmaceutical drugs needed for the treatment of severe human diseases such as cancer, microbial infections and inflammatory processes. Our results evidenced the bioactive potential of the heterotrophic bacterial community of the sponge Erylus discophorus and open the way for further studies.