Phylogenetics and antibacterial properties of exopolysaccharides from marine bacteria isolated from Mauritius seawater

Abstract

Purpose

The marine environment harbours diverse bacterial species which can be exploited for the production of valuable compounds such as exopolysaccharides (EPS) which hold promises for biotechnological applications. The coastal waters of Mauritius is a relatively underexplored marine environment and in this study, isolated bacterial species were tested for the production of EPS exhibiting antibacterial properties against human bacterial pathogens from the genera Acinetobacter, Bacillus, Campylobacter, Enterobacter, Enterococcus, Escherichia, Proteus, Pseudomonas, Salmonella, Streptococcus and Staphylococcus.

Methods

Bacteria were first isolated from seawater samples. Using the disc diffusion method, their EPS were tested for antibacterial effects through two screenings, with each involving a different set of arbitrarily chosen group of pathogens. The microorganisms producing antibacterial EPS were subsequently identified by morphological, biochemical and 16S rRNA-based phylogenetic analyses. Those EPS exhibiting broadest antibacterial activities were eventually characterised by Fourier-transform infrared spectroscopy (FTIR) and thin-layer chromatography (TLC).

Results

Eight EPS were found to display antibacterial effects against more than half of the pathogens and the microorganisms producing them were identified as Bacillus, Halomonas, Psychrobacter and Alcaligenes species. However, only two of these EPS were found to be the most active, with their MIC values ranging between 62.5 and 500 μg/ml. FTIR and TLC analyses revealed the presence of carboxyl, hydroxyl and amide as well as sulphate for the EPS, with glucose or fructose being the main sugar.

Conclusion

The results suggest that Mauritius seawater can be a source of biotechnologically useful microorganisms, producing EPS having potential as antimicrobial agents. DNA sequence data also suggest possible novel bacterial species.

Introduction

Over the past few years, there has been increasing interests in the development of natural polymers, also known as biopolymers, for use in industries as well as for other biotechnological applications (Singha 2012). In this respect, polysaccharides, representing the largest group of produced biopolymers, have often been produced from plants for use mainly as food additives (Vuyst et al. 1998; Poli et al. 2011). However, attention was subsequently shifted to microbial polysaccharides or exopolysaccharides (EPS), the main component of the extracellular polymeric substances. These macromolecules, formed by the polymerisation of sugar residues (Bragadeeswaran et al. 2011; Pawar et al. 2013), are widely diverse in nature with varying molecular weights (10–1000 kDa) and may be classified as either homopolysaccharides, if they consist of only one type of monosaccharide, or heteropolysaccharides, if they consist of multiple ones (Nwodo et al. 2012). Interest in EPS lies mainly in their diversity, with different types of microorganisms producing EPS which vary in terms of composition, size and structure (Yadav et al. 2011). This variation, attributed to differences in the number and types of functional groups attached to the EPS, yields polymers of different properties (Nichols et al. 2005), thereby allowing them to have a wide range of physiological functions within microorganisms such as retention of nutrients to prevent starvation (Singha 2012), in biofilm formation (Vu et al. 2009), for establishing symbiotic relationships (Jones 2012) or for protection against desiccation and presence of toxins or antibiotics in their surroundings (Singha 2012). It is, therefore, not surprising to note that such wide diversity in properties and functions has also made possible numerous industrial applications including in food, cosmetics, textile, pharmaceutical, agricultural, paint, petroleum and waste water treatment among others (Cojoc et al. 2009; Im et al. 2010; Laurienzo 2010).

Currently, the potential applications of EPS in medicine are of particular interest because not only have they been described as useful for medical purposes such as arthritis treatment, wound dressings, surgery or pharmaceutical capsules (Moscovici 2015), but the widely documented health benefits of these compounds, especially from Lactobacillus species, have also prompted investigations on their uses as probiotics (Panthavee et al. 2017; Dinić et al. 2018; Khalil et al. 2018). More recently, biological activities of exopolysaccharides have been reported, with a number of recent studies documenting their immunostimulatory, antiviral, anti-oxidative, anti-tumour and antibacterial properties (Balzaretti et al. 2017; Zhou et al. 2017; Reichert et al. 2017; El-Deeb et al. 2018; Wen et al. 2018). The latter represents an interesting prospect in view of the increasing resistance of pathogenic microorganisms to existing antibiotics (Ghalem 2017). Indeed, drug-resistant microorganisms have become an increasing concern and the potential of microbial exopolysaccharides in fighting microbial resistance lies both in their intrinsic biological activities and their ability to be used in drug delivery systems (El-Naggar et al. 2016; Sezer et al. 2017). Furthermore, EPS with such characteristics may also find applications in food coatings for longer preservation as it is the case for the polysaccharide pullulan (Morsy et al. 2015; Trinetta and Cutter 2016). Hence, screening for new sources of microbial exopolysaccharides may eventually help in developing new antibacterial compounds with different potential applications. The choice of EPS from marine microorganisms is not merely based on the fact that marine microbial natural products, especially those of pharmaceutical value, have become increasingly popular (Debbab et al. 2010). The sea, actually, offers a wider range of biological diversity (Zheng et al. 2011) while at the same time, the aquatic ecosystem, being completely different from the terrestrial one, has caused marine microbes to develop different metabolic and physiological abilities for survival, thereby allowing the production of new compounds not found in terrestrial organisms (Soria-Mercado et al. 2012). As far as microbial extracellular polysaccharides are concerned, the marine environment, therefore, offers greater potential of obtaining new EPS-producing microorganisms as well as novel EPS as demonstrated in some studies (Liu et al. 2013; Gugliandolo et al. 2014; Sun et al. 2015). While it has also been common practice to screen the marine environment for biotechnologically useful microorganisms, such has not been the case of the island of Mauritius. In fact, despite having a very large exclusive economic zone, its marine area has remained largely unexplored with respect to its microbial diversity as well as their potential biotechnological applications. The objectives of this study are as follows: (i) isolate bacterial species from sea water and characterise them based on morphology, biochemistry and DNA sequence data; (ii) extract extracellular polysaccharides from selected bacterial species and assess their antibacterial properties against various human pathogens, including methicillin-resistant Staphylococcus aureus.

Materials and methods

Sample collection

Sampling was carried out during the summer period (November 2014 to April 2015) from three different locations in Mauritius namely Flic en Flac (coordinate 1: 20° 19′ 29″ S; 57° 22′ 8″ E), Tamarin (coordinate 2: 20° 19′ 32″ S; 57° 22′ 33″ E) and La Preneuse (coordinate 3: 20° 21′ 40″ S; 57° 21′ 23″ E). Thirty seawater samples were collected in sterile 125-ml bottles at maximum depth for these sampling sites (between 10 and 15 m) and they were kept on ice until transported to the laboratory where they were stored at 4 °C until use within 24 h. Physical parameters such as temperature, pH and salinity of the water samples were also taken.

Isolation of EPS-producing organisms

Each collected sample was serially diluted in filtered autoclaved seawater and subsequently plated onto four different media namely Plate Count Agar (PCA), Zobell™ Marine Agar (MA) 2216 (HiMedia®), R2A Agar (HiMedia®) and Seawater (SW) Agar, which was autoclaved seawater supplemented with 1.5% (w/v) agar. Plating was carried out in duplicate and the plates were incubated at the same temperature as the collected samples for 24–48 h in the case of MA and PCA while incubation for 3 up to 7 days was carried out for the other two media. In order to obtain additional microorganisms which may be better adapted for growth using seawater as a natural liquid media, samples were also inoculated into filtered autoclaved seawater and incubated for at least 1 week prior to plating onto SW agar. All incubations were carried out not only aerobically but also under anaerobic as well as microaerophilic conditions so that microorganisms with different growth requirements may also be recovered. In this case, plates were kept in airtight jars together with AnaeroGen™ (Oxoid™) and Campygen™ (Oxoid™) sachets for generating anaerobic and microaerophilic atmospheres respectively. Colonies showing visually different characteristics including colour and shape were eventually selected and subcultured on the same medium used as for isolation until pure cultures were obtained. Pure cultures were maintained and stored at 4 °C.

Production and purification of exopolysaccharides by selected organisms

Single colonies of selected isolates were used to inoculate Zobell Marine Broth (HiMedia®) supplemented with 5% (w/v) glucose. After 5-day incubation, exopolysaccharides were extracted based on the method described by Al-Nahas et al. (2011), with modifications. Briefly, culture broth was centrifuged at 5000 rpm for 10 min to pellet cells, after which three volumes of ice-cold absolute ethanol was added to the supernatant before overnight storage at 4 °C to allow precipitation of EPS. The mixture was then centrifuged at 5000 rpm for 20 min to collect the polysaccharides. Purification of EPS was carried out by first dissolving the EPS in a minimum amount of water followed by addition of an equal amount of 20% (w/v) trichloroacetic acid prior to overnight incubation at 4 °C in order to precipitate proteins. After centrifugation at 5000 rpm for 15 min to remove protein pellets, the resulting supernatant was further purified by dialysis (MW cutoff 12,000 Da, Sigma) against distilled water for 48 h at 4 °C. The purified EPS was again precipitated by addition of three volumes of ice-cold absolute ethanol and collected by centrifugation. The resulting pellet was eventually dried to constant mass at 65 °C prior to storage at 4 °C.

Preliminary screening for antimicrobial activity

In order to identify EPS with antibacterial properties, the disc diffusion assay, according to Balouiri et al. (2016), was carried out against seven human pathogenic bacteria—the Gram-positive Bacillus cereus ATCC 10876, Staphylococcus aureus ATCC 29213 and a clinical strain of Staphylococcus saprophyticus as well as the Gram-negative Escherichia coli ATCC 25922, Salmonella typhimurium ATCC 14028, Enterobacter cloacae ATCC 13047 and Proteus mirabilis ATCC 12453. The test strains were inoculated in Muller–Hinton broth and allowed to grow for 24 h at 37 °C. Cultures were then adjusted to a turbidity of 0.5 McFarland (absorbance 0.08–0.1 at 600 nm) and 100 μl of the standardised test strain cultures were plated onto Muller–Hinton agar. Dried EPS from each isolated organism was dissolved in sterile distilled water to a final concentration of 5 mg/ml and 10 μl of each sample was applied to 6-mm sterile discs which were subsequently placed on the agar inoculated with test strains. Plates were incubated in upright position for 18 h at 37 °C. Zones of inhibition around discs were taken as an indication of antimicrobial activity and results for each extract was recorded as the diameter (to the nearest 0.5 mm) of the zone of inhibition produced around each disc. Isolates showing activity against a minimum of four test strains were then selected for further work.

Morphological and biochemical characterisation of selected isolates

Single colonies of the isolates selected from the preliminary screening were first visualised for their general characteristics as described by Smibert and Krieg (1994). This was followed by their biochemical characterisation based on Bergey’s Manual of Determinative Bacteriology (Holt et al. 1994) with tests ranging from the regular Gram-staining for determining bacterial morphology to other biochemical tests including catalase, oxidase, sugar fermentation, urease, sulphite, indole, motility, gelatinase, methyl red-Voges–Proskauer and nitrate reduction.

DNA extraction

DNA was extracted from the selected isolates using the method described by Wilson (2003) with a slight modification. Briefly, 1.5 ml of an overnight-grown culture was centrifuged at 13,000 rpm for 2 min and the resulting pellet was dissolved in 567 μl of TE buffer (10 mM Tris pH 8.0 and 1 mM EDTA pH 8.0). Two microlitres of lysozyme (50 mg/ml) was added and the mixture was incubated for 30 min at 37 °C before addition of 30 μl of 10% (w/v) SDS as well as 3 μl of Proteinase K (20 mg/ml). This was followed by another incubation for 1 h at 37 °C after which 100 μl of 5 M NaCl and 80 μl CTAB/NaCl (10% (w/v) CTAB in 0.7 M NaCl) was added. The mixture was then incubated at 65 °C for 10 min. An equal volume of phenol/chloroform/isoamyl alcohol (25:24:1) was added and the mixture was subsequently centrifuged for 5 min at maximum speed prior to the transfer of the upper layer to a fresh microcentrifuge tube. The same process was repeated a second time using chloroform/isoamyl alcohol (24:1). To the aqueous layer, 0.6 volume of ice-cold isopropanol was added and after 1-h incubation, the mixture was centrifuged at 4 °C. The resulting pellet was then washed with 70% (v/v) ethanol, dried and diluted in TE buffer before being run on 1% (w/v) agarose gel to check for the presence of DNA.

PCR amplification of 16S rRNA and DNA sequencing

The 16S rRNA (1400–1500 bp) was amplified by PCR using the primer pairs 27F (5′-AGAGTTTGATCATGGCTCAG) and 1492R (5′-CGGTTACCTTGTTACGACTT) as forward and reverse primers respectively (Frank et al. 2008). Reactions were carried out in 25-μl reaction volumes using the following components: DreamTaq™ buffer (2.5 mM MgCl2), dNTP (0.05 mM), Primers (0.4 μM) each, DreamTaq™ DNA polymerase (1.5 U) and 40 μg DNA. PCR runs were carried out with the following parameters: initial denaturation for 5 min at 95 °C, followed by 30 cycles, each having a denaturation step at 95 °C (1 min), an annealing step at 55 °C (1 min) and extension step at 72 °C (1 min) before a final extension at 72 °C for 5 min. Amplified products were purified and sent for sequencing with primers mentioned above at Inqaba Biotech South Africa.

Identification of selected isolates and phylogenetic analyses

The 16S rRNA sequences were compared to those of validly published species available in the database of the National Centre for Biotechnology Information (NCBI) using the Basic Local Alignment Search Tool (BLAST). Pairwise nucleotide similarity values between the reference strains and the isolates were determined using the EZBioCloud server (https://www.ezbiocloud.net/) (Yoon et al. 2017). Sequences of closely related species were retrieved and aligned with query sequences using ClustalW (Thompson et al. 1994). Using PAUP 4.0 (Swofford 2002), phylogenetic trees were subsequently constructed using three different tree building methods namely maximum parsimony (Fitch 1971), maximum likelihood (Felsenstein 1981), with the nucleotide substitution model determined using jModelTest (Guindon and Gascuel 2003; Darriba et al. 2012) and eventually neighbour-joining method (Saitou and Nei 1987), for which distance matrices were generated based on Kimura’s two-parameter model (Kimura 1980). The robustness of the tree topologies was evaluated using bootstrap analysis, with 1000 replications of the neighbour-joining method, 500 replications of the maximum parsimony and 100 replications of maximum likelihood.

Secondary screening—antimicrobial activity of EPS from selected isolates and minimum inhibitory concentration

In order to differentiate between the selected EPS for identifying only those exhibiting the broadest range of antibacterial activity, a second set of screening was carried out using only those extracts selected during the previous screening to further compare their biological activity. For this purpose, EPS was extracted as described before and disc diffusion assay was again performed but in this case, against a different set of seven pathogenic strains consisting of Enterococcus faecalis ATCC 29212, Streptococcus pneumonia ATCC 49619, Streptococcus agalactiae ATCC 27956, Campylobacter jejuni NCTC 11351 and three more clinical strains namely methicillin-resistant Staphylococcus aureus, Pseudomonas aeruginosa and a species of Acinetobacter.

This was followed by broth microdilution assay to determine the minimum inhibitory concentration (MIC) of each polysaccharide extract used in the secondary screening against all the 14 pathogenic strains (Balouiri et al. 2016). Basically, 50 μl of Muller–Hinton broth was added to each well of a 96-well microtitre plate. This was followed by the addition of 50 μl of 8 mg/ml EPS to the first well of each column. Fifty microlitres from the first well was then transferred to the second well, leading to a twofold dilution of the EPS and this was repeated up to the eighth well, with frequent pipetting before each transfer for proper mixing. Eventually, from the last one, 50 μl was removed and discarded in order to keep a constant volume in each well. Test strains were cultured as before and were diluted to a concentration of 1 × 10⁶ CFU/ml. Fifty microlitres of that diluted test strain was dispensed into each well to reach a final concentration of 5 × 10⁵ CFU/ml per well and the plates were incubated for 18–20 h under appropriate conditions required for the growth of the pathogenic strains. For each plate, chloramphenicol and Muller–Hinton broth with extract only were used as negative controls while for positive control, inoculated broth without extract was used. After incubation, inhibitory activity of EPS was observed using the method of Eloff (1998). Ten microlitres of p-iodonitrophenyltetrazolium violet (INT)—0.2 mg/ml—was added to each well and the presence of a red coloration indicated the presence of viable cells and hence absence of antimicrobial activity. INT was added to each well of a column until red coloration was obtained for two or more consecutive wells. The MIC value was taken as the minimum concentration of EPS for which no colour change was observed. Results were analysed using two-way ANOVA with test strains and EPS extracts as variables (Minitab 17) to check for differences at 5% significance level.

Chemical characterisation of extracellular polysaccharides

The total carbohydrate content of the polysaccharides with the broadest range of antibacterial activity was spectrophotometrically determined by the phenol-sulphuric acid method (Dubois et al. 1956) using glucose as standard while the presence of protein was determined, using bovine serum albumin as standard, by the Lowry method (Lowry et al. 1951). Sulphate content was also measured as described by Ji et al. (2013), with dipotassium sulphate as standard.

Analysis of monosaccharide composition

The monosaccharide composition of selected exopolysaccharides was qualitatively determined by thin-layer chromatography. The EPS was first hydrolysed, as described by Zhang et al. (2016), with 2 N sulphuric acid for 4 h at 100 °C. The hydrolysed samples were subsequently concentrated prior to application on silica gels (Merck) along with the reference sugars glucose, fructose, rhamnose, ribose, arabinose, xylose and galactose. The chromatogram was developed using acetone/butanol/H₂O (8:2:2) for neutral sugars, butanol/ethanol/H₂O (5:3:2) for oligosaccharides and butanol/H₂O/acetic acid (3:1:1) for acidic sugars (Manca et al. 1996). Visualisation was done by spraying with ethanolic p-anisaldehyde reagent (1 ml p-anisaldehyde, 1 ml conc. sulphuric acid in 20 ml ethanol), α-naphtol-trichloroacetic acid (equal volumes of 0.2% ethanolic α-naphtol + 20% aqueous trichloroacetic acid) and ninhydrin for detection of sugars, uronic acids and amide groups respectively.

Attenuated total reflection Fourier-transform infrared spectroscopy

Attenuated total reflection Fourier-transform infrared spectroscopy (ATR–FTIR) analysis (Bruker Alpha-T) was carried out in order to determine the functional groups present in the exopolysaccharides which were effective against most pathogens. The samples were directly analysed and the spectra were acquired on OPUS Spectroscopy Software (Bruker) in the range 4000–100 cm⁻¹ at 2 cm⁻¹ spectral resolution and with 25 scans.

Results

Sample collection and isolation

For all samples collected, temperature recorded was quite uniform ranging between 27 and 29 °C, with pH being between 8.2 and 8.4. Salinity was also uniform with an average of 3.2%. The seawater showed high microbial load with values between 10⁶ and 10⁷ CFU/ml. This was reflected in the large number of colonies obtained on the different media but eventually, 40 colonies were selected for EPS production based on morphological differences or their mucoid nature.

Preliminary screening

For the preliminary screening, most extracts showed antibacterial activity against at least one test strain, with Figs. 1 and 2 showing the observed zones of inhibition for some of the extracts. Overall, Gram-negative bacteria were more susceptible than the Gram-positive ones. As shown in Table 1, of all the EPS tested, only eight—E5, E15, E16, E18, E19, E20, E26 and E37—(from isolates EPS-5, EPS-15, EPS-16, EPS-18, EPS-19, EPS-20, EPS-26, EPS-37 respectively) were active against a minimum of four pathogenic strains and were, therefore, selected for further screening as a result of their broader range of activity.

Fig. 1

Antibacterial activity of EPS extracts E19, E20, E21, E22 and E23, with red arrow showing diameter of zones of inhibition

Fig. 2

Zones of inhibition produced by E24, E25, E26 and E27, with absence of antibacterial activity for E28

Table 1 Activity of EPS from isolates against test strains for primary screening

Morphological and biochemical characterisation of isolates

Six of the isolates showed similar colony morphology when grown on marine agar, with the colonies being circular, smooth and elevated (Fig. 1) (Supplementary Material), while the remaining two (EPS-20 and EPS-26) were mostly irregular in shape (Fig. 2) (Supplementary Material). Other characteristics of the selected microorganisms were as shown in Table 1 (Supplementary Material) while the source of their isolation, including media on which they have been isolated were as shown in Table 2 (Supplementary Material).

All the isolates were also found to be motile as well as Gram-negative, except for EPS-20 and EPS-26. They were also rod-shaped except for EPS-5 which was coccus-shaped. Further biochemical characterisation showed that all of them were negative for methyl red-Voges–Proskauer tests as well as for indole production. They were also unable to use citrate or gelatin as carbon source with both tests being negative. However, they were all catalase and oxidase positive. Results of the remaining tests for which differences were observed are summarised in Table 3 (Supplementary Material).

DNA-based identification and phylogenetic analyses

Based on 16S rRNA sequences (between 1400 and 1500 bp), it was found that the eight selected isolates could be assigned to two main bacterial phyla namely Proteobacteria and Firmicutes (Figs. 3, 4, 5 and 6). From the phylum Proteobacteria, only EPS-37 belonged to the class Alphaproteobacteria, and closely related to Alcaligenes faecalis (Fig. 3), with which it shared 99.5% nucleotide sequence similarity. Five other isolates from the same phylum could be assigned to the class Gammaproteobacteria and among these, one (EPS-5) was identified as belonging to the genus Psychrobacter, with the isolate having P. ciconiae as its closest phylogenetic relative (Fig. 4). Interestingly, this isolate formed an independent lineage from the P. ciconiae subclade and shared only 96.1% nucleotide similarity with the type strain of P. ciconiae. The remaining four isolates were identified as members of the genus Halomonas. In this case, EPS-15 and EPS-16 formed a distinct clade with H. venusta species while EPS-18 and EPS-19 were more closely related to H. aquamarina and H. meridiana respectively (Fig. 5). To the other phylum Firmicutes, only two isolates, EPS-20 and EPS-26, could be assigned. However, even though they formed a distinct subclade, their phylogenetic relationship with any particular species could not be resolved (Fig. 6).

Fig. 3

Neighbour-joining tree based on 16S rRNA sequences showing the relationship of isolate EPS-37 with other strains from the Alcaligenaceae family. Bootstrap values (expressed as percentages of 1000 replications) ≥ 50 are shown at branch nodes. represent nodes obtained for all phylogenetic trees while represent nodes obtained only on two of the tree topologies. Number in parentheses indicates GenBank nucleotide accession numbers. Oxalobacter vibrioformis WoOx3 was chosen as outgroup. Bar represents 0.007 substitution per nucleotide position

Fig. 4

Neighbour-joining tree based on 16S rRNA sequences showing the relationship of isolate EPS-5 with other strains from the Moraxellaceae family. Bootstrap values (expressed as percentages of 1000 replications) ≥ 50 are shown at branch nodes. represent nodes obtained for all phylogenetic trees while represent nodes obtained only on two of the tree topologies. Number in parentheses indicate GenBank nucleotide accession numbers. Pseudomonas aeruginosa DSM 50071 was chosen as outgroup. Bar represents 0.02 substitution per nucleotide position

Fig. 5

Neighbour-joining tree based on 16S rRNA sequences showing the relationship of isolates EPS-15, EPS-16, EPS-18 and EPS-19 with other strains from the Halomonadaceae family. Bootstrap values (expressed as percentages of 1000 replications) ≥ 50 are shown at branch nodes. represent nodes obtained for all phylogenetic trees while represent nodes obtained only on two of the tree topologies. Number in parentheses indicate GenBank nucleotide accession numbers. Marinomonas mediterranea MMB-1 was chosen as outgroup. Bar represents 0.008 substitution per nucleotide position

Fig. 6

Neighbour-joining tree based on 16S rRNA sequences showing the relationship of isolates EPS-20 and EPS-26 with other strains from the Bacillaceae family. Bootstrap values (expressed as percentages of 1000 replications) ≥ 50 are shown at branch nodes. represent nodes obtained for all phylogenetic trees while represent nodes obtained only on two of the tree topologies. Number in parentheses indicate GenBank nucleotide accession numbers. Staphylococcus saprophyticus ATCC 15305 was chosen as outgroup. Bar represents 0.02 substitution per nucleotide position

Secondary screening—disc diffusion and minimum inhibitory concentration

For the second screening whereby the EPS from the eight isolates were further compared for their antibacterial activity, significant differences were observed between the growth-inhibiting properties of the eight EPS (p < 0.05). As shown in Table 2, only two extracts, namely E15 and E37, inhibited the growth of five and four additional test strains respectively. The remaining polysaccharides were active against less than half of the test strains. The same results were reflected in the MIC assays during which the eight EPS were tested against the whole set of 14 pathogens. In this case, only E15 and E37 inhibited the growth of at least 10 test pathogenic strains (Table 3). The minimum recorded MIC value was 62.5 μg/ml for E37 against Salmonella typhimurium while the rest of the values varied between 125 and 500 μg/ml. As far as the other extracts were concerned, they were inactive against most of the pathogens, except for E26 which was inhibitory to up to seven strains. MIC values, in these cases, ranged from 500 to 2000 μg/ml.

Table 2 Zones of inhibition (mm) produced by each of the eight EPS for secondary screening

Table 3 Minimum inhibitory concentration of selected EPS against pathogenic strains

Chemical characterisation of extracellular polysaccharides and monosaccharide composition

The two selected EPS extracts, E15 and E37, were both made up of around 60% sugar. E15 also had 6% protein content, 22% sulphate content, with fructose being the main sugar detected while E37 had a sulphate content of 19%, a protein content of 7% and consisted of glucose. The presence of amides was also detected in both extracts while the absence of uronic acids could also be noted.

ATR–Fourier-transform infrared spectroscopy

Based on the results of the screenings, E37 and E15 were therefore selected for analysis by ATR–FTIR, with Figs. 7 and 8 showing the spectra obtained. In the case of E37, the broad band at 3272 cm⁻¹ and the less intense ones at 2919 and 2855 cm⁻¹ could be assigned to hydrogen bonded O–H stretching vibrations and to C–H stretching variations respectively, common to polysaccharides (Ordóñez and Rupérez 2011). The signal at 1651 cm⁻¹ could be attributed to a C=O stretch of an acetyl or carbonyl group, with the peak at 1551 cm⁻¹ being attributed to N–H bonding of a secondary amide, which could be linked to protein presence (Chalkiadakis et al. 2013). Among other features, the peaks between 1329 and 1432 cm⁻¹ were likely to represent the symmetric stretching of the carboxylate functional group (COO⁻) or its C–O stretching bands (Kielak et al. 2017; Sardari et al. 2017). There was also the absorption at 1216 cm⁻¹, suggesting an asymmetric stretching of sulphate ester O–S–O (Ordóñez and Rupérez 2011; Yang et al. 2012; Sardari et al. 2017). Finally, for the region from 1200 to 800 cm⁻¹, often considered a fingerprint region used to characterise polysaccharides (Sajna et al. 2013), the strong peak at 1019 cm⁻¹ could be assigned to ring stretching vibrations, C–OH bending as well as the coupled C–O and C–C stretching vibrations, all of which are characteristics of polysaccharides (Choma et al. 2013; Kielak et al. 2017). The peak at 826 cm⁻¹ within this fingerprint region also suggested that E37 could consist of only of α-glycosidic linkages, especially with the absence of any significant signal in the 890 cm⁻¹ region, which would then have been indicative of β-glycosidic linkages (Kalyanasundaram et al. 2012; Du et al. 2017; Kielak et al. 2017). The spectrum for E15 (Fig. 8) had many of the peaks (between regions 500–1600 cm⁻¹ and 3000–4000 cm⁻¹) corresponding to the same functional groups as those present in E37, except for the region 2000–2400 cm⁻¹, which was more likely due to background noise within the spectrum.

Fig. 7

FTIR spectra of E37, showing peaks in the range 4000–400 cm⁻¹

Fig. 8

FTIR spectra of E15, showing peaks in the range 4000–400 cm⁻¹

Discussion

In this work, microorganisms isolated from seawater of Mauritius were screened for their ability to produce polysaccharides which exhibit antibacterial properties. Several authors have reported on the potential of exopolysaccharides as antimicrobial agents (Wilson et al. 2011; Karlapudi et al. 2016; Wu et al. 2016) and this property was demonstrated for EPS obtained from different types of microorganisms (Ghalem 2017; Jeong et al. 2017). Similarly, in this study, eight selected marine isolates were shown to produce EPS which could successfully inhibit the growth of human pathogenic bacteria. Since morphological- and biochemical-based methods were insufficient for their taxonomic classification, identification was largely based on their 16S rRNA sequences through phylogenetic analyses.

Phylogeny generated indicated that EPS-37 could be identified as Alcaligenes faecalis as it clusters together with other A. faecalis strains with strong support (Fig. 3). In fact, our isolate clustered with species obtained from different environments, suggesting that A. faecalis strains, including EPS-37 could be more of a ubiquitous strain rather than a strictly marine one. The other isolate, EPS-5, which could be assigned to the genus Psychrobacter, was more closely related to P. ciconiae. Psychrobacter species are known to be halotolerant (Romanenko et al. 2002) and it could be that our isolate shares similar halotolerance characteristics. However, further investigations are needed to confirm same as Kämpfer et al. (2015) isolated P. ciconiae as a new species from white stork nestling in a non-marine environment. Our phylogeny also depicts EPS-5 as a separate lineage from the P. ciconiae group (Fig. 4) and this isolate shares only 96.1% nucleotide sequence similarity with the type strain. A sequence similarity value less than 98.6% in 16S rRNA with the closest type strain has been suggested to indicate a new species (Kim et al. 2014; Chun et al. 2018) and in fact, for this particular genus, values higher than this threshold have also been reported for novel Psychrobacter species (Zeng et al. 2016). Moreover, EPS-5 does show physical differences from the features of P. ciconiae described by Kämpfer et al. (2015), especially concerning its coccus shape and sugar utilisation. This indicates that EPS-5 could potentially represent a new species but further investigations are warranted.

Four of our selected isolates (EPS-15, EPS-16, EPS-18 and EPS-19) could be assigned to the genus Halomonas. Members of this genus are well known to be halophilic species (Wang et al. 2007) and therefore, can adapt well to saline conditions. In this case, phylogenetic analyses supported the identities of EPS-15 and EPS-16 as H. venusta as well as those of EPS-18 and EPS-19 as H. aquamarina and H. meridiana respectively (Fig. 5). On the other hand, the identification of isolates EPS-20 and EPS-26 was limited to genus level. They clustered along with other Bacillus species (e.g. B. paramycoides, B. anthracis, B. cereus, B. thuringiensis) but this subclade (Fig. 6) received moderate support and phylogenetic relationships with any particular species within the clade could not be resolved (sequence variation was very minimal). In fact, there has been concerns on the reliability of 16S rRNA sequences to delineate species with high level of nucleotide similarities (Janda and Abbott 2007; Woo et al. 2009) and other genes exhibiting more sequence variability are needed. This applies particularly for species such as Bacillus cereus and its close relatives whereby classification is difficult given that high similarity was obvious even when whole genomes were compared (Schmidt et al. 2011; Okinaka and Keim 2016). This led to the suggestion that they could be genetically classified as the same organism (Zahner et al. 2013). However, even though DNA–DNA hybridisation values less than the 70% threshold indicated that they were indeed distinct organisms (Guinebretière et al. 2013), the species of this subclade are often referred to as the Bacillus cereus group (Ceuppens et al. 2013; Zahner et al. 2013) but to avoid nomenclatural confusion, we label our isolate as an unidentified Bacillus species.

In this study, several isolated marine microorganisms were shown to produce EPS which could successfully inhibit the growth of human pathogenic bacteria, with Gram-negative bacteria being more susceptible than the Gram-positive ones, except for Staphylococcus aureus which was also very susceptible. These two types of bacteria differ in their cell walls, with Gram-negative bacteria having a thinner one as well as an outer layer compared to Gram-positive bacteria (Silhavy et al. 2010; Brown et al. 2015). Hence, the observed differences in susceptibility could very likely be due to the way in which the EPS interact with the different cell walls. In fact, the efficiency of other antimicrobial compounds has also been linked with differences in the cell walls between the two types of bacteria (Prochnow et al. 2016). This includes the polysaccharide chitosan, for which its positively charged structure and hydrophilic nature is considered essential for antimicrobial activity against Gram-negative bacteria (Kong et al. 2010) since the thin surfaces of these organisms are negatively charged, resulting in electrostatic interactions which may not only disrupt cell wall permeability but also cause its hydrolysis and subsequent leakage of cell contents (Goy et al. 2009). However, it should be noted that, as far as chitosan is concerned, there are conflicting reports on which group of bacteria is more susceptible (Kong et al. 2010). In addition to this, the growth of Gram-positive organisms such as Staphylococcus aureus was also inhibited while no effect on Escherichia coli, a Gram-negative one, was observed. These suggest that the mechanism of action of the isolated exopolysaccharides could be more complex than simple interaction with the bacterial cell walls.

The secondary screening as well as the MIC assays allowed the EPS from the eight isolates to be further differentiated based on their antibacterial activity and eventually polysaccharide extracts E15 and E37, from isolates EPS-15 (Halomonas sp.) and EPS-37 (Alcaligenes faecalis) respectively, were found to be particularly effective against most pathogens. Interestingly, antibacterial EPS of organisms from these two genera is not widely documented and both sets of screenings indicated that polysaccharides from Halomonas and Alcaligenes species might be promising antimicrobial agents against various pathogenic strains. In fact, one of their most promising features was the inhibitory effect exerted on Staphylococcus aureus, including methicillin-resistant Staphylococcus aureus (MRSA). The latter is a major cause of nosocomial infections (Dakheel et al. 2016) and its incidence has been on the rise, becoming a growing public health concern, especially due to its resistance to a wide range of antibiotic treatment (Sardi et al. 2017; Neopane et al. 2018). Although antimicrobial activity of EPS has been studied against different pathogenic strains, their effectiveness against MRSA has not been widely investigated. These results may, therefore, help to further study the different ways in which antimicrobial polymers such as EPS can be further used, along with known drugs, to develop more effective treatments against pathogenic microorganisms, especially antibiotic-resistant ones.

The use of FTIR allowed the determination of functional groups present in E15 and E37 for identifying those which could be linked to the observed bioactivity. The spectra for both extracts did not show major differences, with the presence of groups such as carbonyl, amine or even hydroxyl being observed in both. These functional groups have the ability to chelate metal ions (Collin et al. 2015) and this property is more often exploited for bioremediation purposes (Philippis et al. 2011; Gupta and Diwan 2017). However, the ability to sequestrate metal ions has also been suggested as a potential mechanism for antimicrobial activity, including for the polysaccharide chitosan (Sebat et al. 2001; Kong et al. 2010; Qiu et al. 2011). Indeed, metal ions are essential as nutrients for microbial growth and at the same time, are required for the stability of microbial cell walls by combining with molecules within the walls (Goy et al. 2009; Kong et al. 2010). Hence, through the chelation process, the availability of metal ions to microorganisms are reduced and this could have also contributed to the observed antibacterial properties.

The FTIR results as well as the compositional analysis of E37 and E15 also showed that they bear some similarity with the polysaccharides curdlan and levan respectively. This is likely given that these two polymers can be produced by Alcaligenes and Halomonas species, the same genera to which the isolates producing E37 and E15 belong. Interestingly, although the antibacterial properties of curdlan and levan is known (Byun et al. 2014; Chen and Liang 2017), their effectiveness against a wide range of microbial pathogens, as carried out in this study, has not been investigated. Hence, even though further structural characterisation is required to assess the extent to which E15 and E37 is similar to curdlan and levan, yet the observed antibacterial properties in this work suggests that both curdlan and levan might also exhibit biological activities against a wide array of pathogens. Furthermore, it was interesting to note the presence of sulphate groups in the two extracts since sulphate-containing polysaccharides of marine origin are quite important for industries (Ladrat et al. 2014). In fact, these have been attracting attention due to a number of their unique physico-chemical properties (Raveendran et al. 2013) as well as a number of their biological properties (Jesus Raposo et al. 2013; El Essawy et al. 2016; Heymann et al. 2016), thereby further highlighting the potential of the two isolated exopolysaccharides as bioactive compounds.

Conclusion

The search for new compounds of biotechnological value has greatly fuelled the need to probe new environments. In this work, Mauritius seawater, a relatively underexplored one, has been shown to harbour different microbial species, especially those which can produce antibacterial exopolysaccharides. Among these exopolysaccharides, two sulphated ones seemed promising as inhibitors of the growth of human bacterial pathogens, including antibiotic-resistant ones. Although further work is required to assess their suitability as antibacterial agents, this work highlights the potential of using the isolated polymers in the development of new drugs or more effective treatments. Furthermore, this study shows that Mauritius seawater can be a source of biotechnologically useful microorganisms producing compounds of potential value.