Screening, optimization and characterization of exopolysaccharides produced by novel strains isolated from Moroccan raw donkey milk.

Two exopolysaccharides (EPS) producing strains, isolated from raw donkey milk were identified as Leuconostoc mesenteroides SL and Enterococcus viikkiensis N5 using 16S rDNA sequencing. The Box Benheken design exhibited the highest yield of EPS-SL (672.342 mg/L) produced by SL and of EPS-N5 (901 mg/L) produced by N5. The molecular weight was 1.68×104 for EPS-SL and 1.55×104 Da for EPS-N5. FTIR, NMR and GC-MS analysis showed that the EPS are heteropolysaccharides. The SEM image showed that the EPS-SL was smooth and represented a lotus leaf shape and EPS-N5 revealed a stiff-like, porous appearance and was more compact than EPS-SL. The TGA analyses showed high thermal stability and degradation temperature. Additionally, the two EPSs possessed antibacterial and antioxidant activity, and the EPS-SL had the stronger antioxidant activity. Consequently, these results suggest that the functional and biological properties of EPS-SL and EPS-N5 imply the potential application in the food and pharmaceutical industries.

Introduction

Because of its persuasive nutrition and functional ingredients, donkey milk has piqued the interest of academics in recent years (Li et al., 2020a). Its chemical composition is very similar to that of human milk due to content in major components as well of minor components (like oligosaccharides)(Derdak et al., 2020). As a result, it has been utilized as an acceptable alternative for babies who have a variety of food allergies and intolerances (Derdak et al., 2020, Vincenzetti et al., 2017). Fermented donkey milk products have recently been presented as essential sources of probiotics and antioxidants with a variety of health advantages. It has increased in popularity as a result of its biological properties, such as antibacterial and antioxidant activity (Vincenzetti et al., 2017). Its antibacterial action resulted from a large microbial variety, primarily lactic acid bacteria (LAB), which are widely regarded as safe microorganisms (GRAS) because of their long and safe usage as starter cultures in fermented foods (Kostelac et al., 2021). LABs are bacilli and cocci-shaped bacteria that form a wide range of microorganisms. The production of lactic acid, which is the end result of the fermentation of various sugars, is a common trait of these bacteria (Derdak et al., 2021, Sakoui et al., 2020). Furthermore, these microorganisms may be a source of antimicrobial and antifungal metabolites such as organic acids, reuterin, hydrogenperoxide, diacetyl, carbondioxide, bacteriocins, and exopolysaccharides (EPS), which inhibit pathogens contaminating milk during processing, transport, storage, and preparation (Daba et al., 2021, Kostelac et al., 2021).

EPSs produced by lactic acid bacteria (LAB) are the subject of an increasing number of studies. LABs can produce EPSs that are potentially useful as safe additives to improve texture and viscosity of natural fermented milk products and as stabilizing, emulsifying, and antioxidant agent (Jiang et al., 2021). Moreover, it has been suggested these biopolymers may confer health benefits to the consumer. Some studies have indicated that these EPSs may have immunostimulatory and anti-tumoral activity (Wu et al., 2021).

These exopolysaccharides are generally composed of repeated units of oligosaccharides. They are divided into two groups: homopolysaccharides and heteropolysaccharides, depending on whether they are made up of one or more types of sugar, and the qualities listed above are all linked to the chemical structure of these substances (Wang et al., 2015a). In consequence, the determination of these carbohydrate linkages is of great importance. Jolly et al. reported that sugar composition, molecular weight, sugar linkages, presence of repeated units and substitutions, affect the biological and technological properties of EPS. Thus, the chemical characterization of novel EPS is crucial for predicting their potential application (Jolly et al., 2002).

To our Knowledge, no study reports the evaluating and characterization of EPSs produced by strains isolated from donkey milk. Therefore, this study was conducted to optimize the production of EPSs from two lactic acid bacteria newly isolated from donkey milk, characterize them and evaluate their antibacterial and antioxidant activity. EPS production was optimized by the Response Surface Methodology (RSM). The characterization of those EPSs was performed using the scanning electron microscopy (SEM), the Fourier transform infrared (FTIR), the Nuclear magnetic resonance (NMR) spectroscopies and the Gas Chromatography Coupled with Mass Spectrometry (GC–MS). Moreover, the thermic stability of the studied EPSs was evaluated by thermogravimetric analysis (TGA) and molecular weight was determined using an U-shaped Ostwald viscometer. Then, the antioxidant and antibacterial activities were also reported.

Materials and methods

Chemicals and reagents

MRS, yeast extract, tryptone, peptone, Mueller-Hinton broth, potassium persulfate were obtained from Biokar diagnostics (Beauvais, France). Sucrose, glucose, lactose, d-glucose, d-xylose, l-arabinose, d-galactose, d-mannose, trichloracetic acid (TCA), potassium bromide (KBr), trifluoroacetic acid, ammonium molybdate, resazurin, pyridine, methanol, ammonium hydroxide, ascorbic acid, acetic anhydride and glacial acetic acid were procured from Sigma-Aldrich (Saint louis, MO, USA). 1,1-diphenyl-2-picrylhydrazyl (DPPH) and 2,2-azinobis-(3-ethylbenzothiazoline-6-sulphonic acid) (ABTS) were purchased from Fluka (Basel, Switzerland). Aluminum chloride (AlCl3), potassium ferricyanide (K3Fe(CN)6), sodium borohydride, phenol and chloroform were obtained from Merck (Darmstadt, Germany).

Sampling and isolation of strains

A sample of donkey's milk was obtained in a sterile tube and transported at 4 °C. The isolation of LAB was carried out on MRS medium agar.

Screening of EPS-producing strains

Isolated strains were screened for EPS-producing phenotype using different methods: i) detection of “ropy” “phenotype (formation of an unbreakable strand when the colony is touched with an inoculation loop)” and “mucoide non-ropy” “phenotype (glistening, smooth and slimy appearance without occurrence of filament) of colonies growing in MRS media and MRS supplemented with 2% of sucrose; ii) viscosity of liquid cultures using” “exopolysaccharide selection medium” “ (ESM: 90 g skimmed milk, 3.5 g yeast extract, 3.5 g peptone and 10 g sucrose per litre)”.

Molecular identification

A chloroform phenol technique was used to extract DNA, as previously described (Derdak et al., 2021). Total genomic DNA was isolated from samples and utilized as a template for polymerase chain reaction amplification of a 16S ribosomal DNA fragment (PCR). The PCR reactions have been conducted through a DLAB TC1000-G thermocycler. Two pairs of conserved universal primers were used for amplification: Fd1 5′-AGAGTTTGATCCTGGCTCAG-3 'and RP2 5′-ACGGCTACCTTGTTACGACTT-3′.

The PCR was performed in a total volume of 15 µL composed of approximately 100 ng of genomic DNA, 0.3 μL of each 10 mM dNTP (Promega), 3 μL of 5X buffer (Promega), 0.9 μL of 25 mM MgCl2 (Promega), 0.12 μL of each 10 mM primer and 0.075 µL of Taq polymerase 5U (Promega). The PCR program used was as follow:

95 °C, 2 '/(95 °C, 40′ '- 55 °C, 40′ '- 72 °C, 1′) x35; 72 °C/5 '/4 °C.

The ExoSAP-IT purification kit (GE Healthcare) was used to purify the PCR products, and the BigDye Terminator Kit version 1.0 was used to sequence them (Applied Biosystems). The sequences obtained were corrected using BioEdit, introduced into Geneious prime and aligned via BLAST to determine the degree of similarity of our sequences with those present in the database.

Response surface optimization experiment

Significant factor determination Single-factor

The influence of different parameters was studied by growing the strain in modified (mMRS) or unmodified MRS. For screening carbon source, glucose present in MRS medium was replaced by sucrose or lactose while keeping other components in the medium constant. For screening the nitrogen source, peptone, yeast peptone, or tryptone were used while keeping other components constant. While the inoculum percentage, incubation temperature and incubation time, were constant in all the above conditions, as 1%, 30 °C and 24 h, respectively. The above carbon and nitrogen sources have concentrations of 2%.

For identifying the optimum conditions of inoculum percentage, incubation temperature (30–37 °C) and incubation time (18-48H), unmodified MRS was used. The quantity of total sugar was evaluated using the phenol-sulfuric acid technique (Dubois et al., 1956).

Response surface optimization experiment

Sucrose, peptone, and % inoculum all had a substantial influence on the EPS, according to single-factor screening results. As a result, these three variables were chosen for response surface design (RSM). Design-Expert 11 was used to create the experimental matrix (Table S1).

A Box Behnken design was used to collect experimental data that fits in a complete Quadratic polynomial model describing the response surface across a relatively large range of parameters. Tables S2 and S3 show the range and levels of the experimental variables evaluated. The quadratic equation (Eq. (1)):Where Y is the measured response, A, B and C are the coded independent input variables, is the intercept term,,, and are the coefficients showing the linear effects,, and, are the quadratic coefficients showing the squared effects and, and are the cross-product coefficients showing the interaction effects. The experiment was repeated 17 times, and the data from the design experiment was utilized to do regression analysis using the program.

Isolation and purification of EPS

The strains were cultivated in MRS medium in optimum conditions found previously. After incubation, the cell-free supernatant was collected by centrifugation at 10000× g for 10 min, and then mixed with 5% trichloroacetic acid (TCA) and kept at 4 °C for 30 min. Centrifugation at 10000× g for 10 min at 4 °C was used to remove the proteins. The supernatant was diluted with two volumes of cold ethanol and kept at 4 °C overnight. The precipitated crude EPS was then recovered using a 10000× g centrifuge at 4 °C for 15 min. The EPS was dissolved in deionized water and dialyzed against deionized water for 48 h at 4 °C using dialysis tube with a MW cut-off of 10 kDa, before being freeze-dried. The crude EPS was fractionated on a Sephadex G-100 column (1.6 cm × 50 cm) and eluted with distilled water at a flow rate of 1 mL/min for further purification. The carbohydrate content of every 4 mL of elution solution was measured using the phenol-sulfuric acid technique (Dubois et al., 1956). The peak fractions were combined, dialyzed, and frozen. For further examination, the pure EPS was employed.

Molecular weight

Different concentrations of samples were produced, and in a U-shaped Ostwald viscometer, the flow time of equal volumes for each concentration at 30 °C was calculated (Gamal et al., 2021). The flow time of the same volume of distilled water was also measured as a control. Thus, specific viscosity/C was computed by converting the relative viscosity value using the formulae below; the specific activity was derived by converting the relative viscosity value using the equations below. (Eqs. (2) and (3)):Where ŋr is the relative viscosity, and ŋsp is the specific activity. t0, t and α0, α are the time of flux, and the density of samples and solvent, respectively.

The value of the intrinsic viscosity [ŋ] was determined by plot [(ŋ-ŋ0)/ŋ0c), where C is the concentration of the sample. It is also defined as the limit of ŋsp/C when the concentration approach to 0 (Eq. (4)):

The molecular weight was calculated using Mark-Houwink-Sakurada equation (Eq. (5)) (Gamal et al., 2021):

Gas Chromatography Coupled with Mass Spectrometry (GC–MS)

The composition of EPS was detected by the method described by Fooladi et al 2019 with slight modifications (Fooladi et al., 2019). 20 mg of EPS was hydrolysed with 1 mL of 2 M trifluoroacetic acid at 100 °C for 2 h. Methanol was added to the sample after hydrolysis and evaporated. The hydrolysate was then converted to alditol acetates using 1 mL sodium borohydride (10% in 1 M ammonium hydroxide) for 1 h at 30 °C. After that, the samples were neutralized with glacial acetic acid and evaporated. 1 mL acetic anhydride and 0.5 mL pyridine were used to acetylate the reduced sugars, which were then incubated at 50 °C for 40 min. The solution was then extracted and evaporated using chloroform. The mixture was reconstituted with 1 mL of methanol before being analyzed using GC–MS (Shimadzu QP-2010 plus) on a capillary column BP-5 (30 m × 0.25 mm × 0.25 mm).

The monosaccharide was detected using the GC–MS library and validated using the standard monosaccharide (d-glucose, d-xylose, l-arabinose, d-galactose, and d-mannose), which had been derivatized using the same processes as previously indicated.

Fourier-transform infrared spectroscopy (FTIR) and ultra-violet spectral (UV) analysis.

To investigate alternative functionalities, an IRAffinity-1S spectrometer was used to gather FTIR for the EPS-SL and EPS-N5. 2 mg of samples were mixed with 200 mg of KBr to make compressed discs with a diameter of 3 mm, and the spectra was adjusted for KBr background. The pellets were then scanned at a resolution of 4 cm−1 and 32 scans in the range of 4000–500 cm−1.

2 mg of samples was dissolved in 1mL of distilled water and analyzed using an UV–visible spectrophotometer, and the spectra were recorded between 190 - 400 nm (Shimadzu UV-1800).

Nuclear magnetic resonance (NMR)

The EPS-SL and EPS-N5 samples were solubilized in deuterium oxide (D2O) and was analyzed. 1H and 13C NMR spectra were registered on a JNM-ECZ500R/S1 FT NMR SYSTEM (JEOL) 500 MHz spectrometer.

Scanning electron microscopy (SEM)

SEM (QUATTRO S-FEG -Thermofisher scientific) was used to examine the surface morphology and microstructure of EPS-SL and EPS-N5 at magnifications of 1000 and 2500. The lyophilized EPS (20 mg) was mounted on a SEM stub, gold sputtered, and analyzed by SEM.

Thermogravimetric analysis

The thermal analysis (TGA) was recorded by a thermogravimetric DTG-60H (SHIMADZU). The EPS-SL and EPS-N5 were placed in an Al2O3 crucible and heated from 10 °C to 600 °C at 10 °C/min. The experiment was performed under a nitrogen atmosphere.

Antioxidant activity

DPPH radical scavenging activity

The DPPH radical scavenging activity of the EPS-SL and EPS-N5 was measured by the Adesulu-Dahunsi method with modifications (Adesulu-Dahunsi et al., 2018). Different concentration of EPS-SL and EPS-N5 (1, 2,3 and 5 mg/mL) was prepared. 1.5 mL of each concentration was mixed with 1.5 mL of DPPH (2.3 mg in 100 mL of methanol).

The mixture was vortexed forcefully and held at room temperature for 30 min before being measured at 517 nm.

The positive standard was ascorbic acid. The experiment was carried out three times.Where is the absorbance of blank and , the absorbance of the sample.

ABTS radical scavenging activity

The ABTS radical scavenging activity was carried out using the Bai technique with minimal changes (Bai et al., 2021). The ABTS radical solution (a combination of 7 mM ABTS and 2.45 mM potassium persulfate) was diluted with PBS to an absorbance of 0.70 ± 0.02 at 734 nm (pH 7.4). 1.5 mL of each EPS sample was combined with 2.5 mL of ABTS radical solution (concentration 1–5 mg/mL). The combination was then reacted at room temperature for 6 min before being measured at 734 nm. Each experiment was carried out three times in total.

As a positive control, ascorbic acid was employed. The radical scavenging activity of ABTS was determined using the formula (Eq (7)):Where is the absorbance of blank and , the absorbance of the sample.

Reducing power

The reducing power was calculated in accordance with Amiri et al, with minor modifications (Amiri et al., 2019). Briefly, 1 mL sample at various concentrations was mixed with 0.2 mol/L PBS (2.5 mL) and 1% K3[ Fe (CN)6] solution (2.5 mL). After incubation at 50 °C for 20 min, 10% TCA (1 mL) was added and centrifuged at 3000× g for 10 min. Then, the upper layer (1 mL) was combined with distilled water (0.6 mL) and 1% FeCl3 (0.4 mL). Finally, the absorbance was determined at 700 nm. Each experiment was carried out three times with ascorbic acid as a positive control.

Phosphomolybdenum assay

The antioxidant activity of EPS-SL and EPS-N5 was evaluated by phosphomolybdenum method (Khan et al., 2012). 1 mL of reagent solution was mixed with an aliquot of each concentration (0.6 M sulphuric acid, 28 mM sodium phosphate and 4 mM ammonium molybdate). The samples were cooled to room temperature and the absorbance was measured at 765 nm after 90 min in a water bath at 95 °C. Each experiment was carried out three times in total.

Antibacterial activity

Bacteria and culture conditions.

Five strains were used for this study, three Gram-positive bacteria; Staphylococcus aureus, Listeria monocytogenes and Streptococcus pyogenes, and two Gram-negative bacteria; Escherichia coli and Salmonella thyphimurium. All the strains were cultured in Mueller-Hinton and incubated at 37 °C for 24 h.

Determination of minimum inhibitory concentrations (MIC)

Antimicrobial activity was assessed using a modified microdilution approach. A serial dilution procedure was used to determine the MIC using 96 well plates. With wells containing 100 µL of Mueller-Hinton broth, different dilutions of the EPS-SL and EPS-N5 were carried out, and then 10 µL of inoculum was introduced to all of the wells. The microplates were incubated for 24 h at 37°Celsius. After adding 20 µL of resazurin (200 µg/mL) solution to each well and incubating the plates for 2–3 h at 37 °C, the MIC of the samples was determined.

A switch from blue to pink suggests a decrease in resazurin levels and, as a result, bacterial proliferation.

Statistical analysis

Analysis of variance (ANOVA) was used to examine the results, and the Tukey test was used to detect significant differences using GraphPad Prism 7.0 software.

Results and discussion

Screening and identification of EPS-producing strains

10 LABs were isolated from Moroccan raw donkey milk. These LABs were screened on MRS supplemented with sucrose. 8 isolates showed no viscous colonies and 2 isolates showed a viscous and ropy morphology. The 2 LAB isolates were subjected to molecular identification. The 16S rDNA was used as query sequences to BLAST in Geneious Prime Software, and the sequences with high pairwise identity were selected. The strains were identified as Leuconostoc mesenteroides SL and Enterococcus viikkiensis N5.

Table S1 (Supplementary material) shows the impact of various carbon and nitrogen sources, as well as inoculum variables, on EPS generation by LS and N5. The three carbon sources had a statistically significant difference (p < 0.05), with sucrose (503 and 712 mg/L) being more advantageous for EPS generation by SL and N5, respectively. Our results were consistent with those obtained by Vidhyalakshmi et al. (2016). The findings of the nitrogen source screening revealed that peptone, at levels of 501 and 698 mg/L, was more suitable for the synthesis of EPS by SL and N5. The carbon and nitrogen sources in the culture medium have an impact on EPS generation (Imran et al., 2016).

All those conditions were done with 1% of inoculum, and the change of temperature and incubation time doesn’t show any differences in the EPS production. When increasing the percentage of inoculum to 2%, the amount of EPS increased to 495 mg/L. Among the different conditions, the inoculum was proved to be more beneficial, therefore was chosen for further optimization. Verma et al showed that the inoculum influences the production of EPS (Verma & Kumar, 2020). Our results were close to those found by Majumder et al. (2009). Additionally, the ability of N5 to generate EPS was higher than EPS produced by SL.

The RSM studied the effect of 3 variables (Sucrose, peptone and inoculum) on EPS production. Tables S2 and S3 show the Box-Behnken design matrix for EPS generation by SL and N5, respectively, with experimental, actual, and anticipated values. There were 17 experimental runs, each in triplicate, based on the model developed by the program. The equation used to forecast the highest EPS output is shown below after fitting the experimental findings into the Box-Behnken design under the response surface optimization experiment (Eq. (8) and (9)):

Where A is sucrose, B is peptone and C is inoculum.

The analysis of variance on RSM to optimize EPS production by SL and N5 are shown in Table S4. The results showed that the ANOVA for model 1 (model of EPS produced by SL) and model 2 (model of EPS produced by N5) was significant (p < 0.0001). The significance of Models 1 and 2 was further validated with F-values of 2381.69 and 80.15, respectively.

Model 1 showed that sucrose, peptone and inoculum, and their interactions had statistically significant effects on EPS produced by SL (p < 0.0001). However, model 2 showed sucrose and peptone and their interaction had statistically significant effects on EPS produced by N5 (p < 0.005). Despite this, the concentration of inoculum had no statistically significant influence on EPS production (p < 0.05).

Model 1 had a correlation value of 99.93 % and model 2 had a correlation coefficient of 99.04 %, showing that the regression of the two models was good. Similarly, the corrective decision coefficient (RAdj2) for models 1 and 2 was 99.89 % and 97.8 %, respectively, indicating that most test data variability can be explained by the models. Two independent variables were used to create the three-dimensional response surface graphs, which were used to investigate the ideal level of each variable and its impact on EPS generation.

The latter assists in comprehending the link between each variable's response and experimental levels. These graphs also indicate the interaction between test factors, which aids in determining the best condition (Vidhyalakshmi et al., 2018). By evaluating all feasible combinations, Fig. 1 showed a total of three response surfaces for EPS created by SL and N5, respectively.

Fig. 1

3D plot for interactions between independent variables on EPS production by SL (A– C) and N5 (D-F).

The interaction between the tested factors is depicted in these charts, which allows us to find the best condition (Vidhyalakshmi et al., 2016). The maximum amount of EPS-SL (672.34 mg/L) was obtained by a combination of 6% sucrose, 1% inoculum and 2.5% peptone, and for EPS-N5, 901 mg/L was produced by a combination of 6% sucrose, 2.5% inoculum and 4% peptone. While Amiri et al found that the maximum amount of EPS produced by Bifidobacterium animalis subsp. lactis BB12 was 160,781 mg/L (Amiri et al., 2019).

According to Fig. 1(A and B), the yield of EPS-SL increased by increasing sucrose and decreasing peptone and inoculum, and there was a noticeable interaction between inoculum and peptone (Fig. 1C). Furthermore, the yield of EPS-N5 increased by increasing sucrose, peptone and inoculum (Fig. 1D–F).

Purification and molecular weight determination

The crude EPS extracted from SL and N5 was purified by Sephadex G-100 column (Fig. S1; Supplementary material). The results showed a single peak for each EPS, indicating homogeneous polysaccharide. This peak was eluted and freeze-dried. EPS-SL and EPS-N5 have average molecular weights of 1.68×104 and 1.55×104 Da, respectively. Lactic acid bacteria generate EPS with a molecular weight ranging from 104 to 106 Da (Daba et al., 2021).

The GC–MS was performed to analyse the monosaccharide composition of EPS-SL and EPS-N5 after alditol acetate derivatization. The results are represented in Fig. 2.

Fig. 2

GC–MS chromatograms of A: EPS-SL, B: EPS-N5 and C: Standards.

In comparison with standards, the results showed that EPS-SL contained d-glucose (62.84%) and d-galactose (37.16%). However, EPS-N5 was mainly composed of d-glucose (79%), d-galactose (12.7%) and d-mannose (8.3%). These results suggest that the EPS-SL and EPS-N5 were heteropolysaccharides.

The EPS generated by Leuconostoc mesenteroides XR1 contained glucose and galactose (L. Wang et al., 2021). Li et al presents that the EPS produced by Leuconostoc mesenteroides SN-8 was composed of glucan and mannose (Li et al., 2020b). The EPS produced by Lactobacillus acidophilus LA5 was composed of glucose (23.25%), mannose (24.49%), galactose (18.46%), and glucuronic acid (19.50%), xylose (8.45%) and fructose (5.85%) (Amiri et al., 2019). Wang et al 2010 showed that EPS produced by Lactobacillus plantarum KF5 was composed of glucose, mannose and galactose (Y. Wang et al., 2010).

Fourier-transform infrared spectroscopy (FTIR) and ultra-violet spectral (UV) analysis

The UV spectra of the EPS-SL and EPS-N5 revealed no absorption peak at 260 nm or 280 nm, indicating the no existing of aromatic amino acid and nucleic acid (data not shown).

FTIR examination of bacterial EPS revealed a number of peaks that are typical of carbonyl compounds (Fig. 3). The FTIR spectra of the EPS-SL and EPS-N5 samples are presented in the Fig. 3. The two samples (which had comparable spectra) displayed a broad and bright band at roughly 3340 cm−1, which was attributed to the stretching vibration of the carbohydrate hydroxyl groups (Hu et al., 2019), and also a C—H stretching band at around 2930 cm−1. These absorbances represent the polysaccharide fingerprint peak (Amiri et al., 2019). The region of 1655 cm-1 corresponds to the stretching vibration of C O.

Fig. 3

Fourier-transformed infrared (FT-IR) spectrum of A: EPS-SL and B: EPS-N5.

In addition, a band at 1400 is due to C—O, C O (Cao et al., 2020). Both samples had a specific absorption at 1000–1250 cm−1 attributed to the presence of C—O—C glycosidic bonds and C—O—H link bonds of the pyranose unit in each polysaccharide (Ayyash et al., 2020). The absorption peak at 843.59 cm−1 was generally formed by α-glycosidic linkages, while the peaks at approximately 915 cm−1 were diagnostic of pyranose (Cao et al., 2020). The spectrum showed no significant difference. Our findings matched those of Cao et al and Vidhyalakshmi et al (Cao et al., 2020, Vidhyalakshmi et al., 2016).

NMR spectroscopy is one of the most often utilized characterization techniques for EPS structural analysis. Fig. 4 shows the 1H and 13C NMR spectra of the EPS-SL and EPS-N5. Generally, the 1H NMR is used to analyze the glycosidic bond configuration of polysaccharides. The proton signal of α-anomeric pyranose is >4.8 ppm, and the proton signal of β-anomeric pyranose is <4.8 ppm (Cao et al., 2020). The anomeric zone (δ 4.5–5.5 ppm), where signals of the anomeric protons of each sugar residue are frequently presented; the ring protons region (δ 3.1–4.5 ppm); and the alkyl groups region, which ranges between δ 1.2–2.3 ppm, make up the 1H NMR spectra of polysaccharides. In the present study, the 1H NMR spectra, two anomeric signal were observed for EPS-SL at δ 4.82 and at δ 4.9, and three signals were observed for EPS-N5 at δ 4.85, δ 4.9 and δ 5.1. These peaks were characteristics signals of α- (1 → 6) and α- (1 → 3) glycosidic linkage (Bounaix et al., 2009; J. Li et al., 2021). The absorption peak at 4.66 ppm was attributed to D2O.

Fig. 4

NMR spectrum of the purified EPS produced by SL: 1H (A) 13C (B) and N5 1H (C) 13C (D).

The 13C NMR spectrum (Fig. 4) of EPS-SL and EPS-N5 included anomeric carbons (95–110 ppm) regions and ring carbons regions (50–85) (Ye et al., 2019). Since the α-pyranose alpha carbon signal is typically δ 97–101 ppm, and the β-pyranose residue is typically δ 101–105 ppm (Cao et al., 2020). Two anomeric carbons regions at δ 97.66 and 97.71 for EPS-SL and 97.78 and 97.85 for EPS-N5 corresponded to α linkage. The other sugar carbon signals appeared between 98 and 65 ppm in the 13C NMR spectrum. A similar kind of 13C NMR interpretation of EPS was reported through signals at 60 to 70 ppm revealed glucose and galactose residues (Wang et al., 2015a). There was no peak at δ >170 ppm, indicating that uronic acid was not present.

The TGA of the EPS-SL and EPS-N5 was done dynamically between weight loss and temperature changes (Fig. S2). The results of the TGA analysis showed three different phases. The EPS-SL and EPS-N5 had an initial weight loss of around 12% from 10 to 150 °C. This immediate decrease of mass might be related to moisture loss. For the EPS-SL no significant mass changes were observed between 150 and 280 °C, which was the highest temperature that EPS-SL can resist. Subsequently, a sharp weight loss (73.5%) was observed from 280 to 340 °C. However, for the EPS-N5, a weight loss of 44.8% was observed between 150 and 340 °C.

The EPS mass then steadily dropped, leaving a final residue of 4.5% for EPS-SL and 15.9% for EPS-N5, respectively. Varied carbohydrate compositions are likely to blame for the EPSs' different temperature stability and degradation behavior (Wang et al., 2015b). Thermal stability is a crucial feature to consider for industrial applications of EPS, notably in the food sector, because most food items are manufactured and processed at high temperatures (Bomfim et al., 2020). The thermostable EPS-SL and EPS-N5 are excellent choices for food products.

The scanning electron microscope (SEM) is a vital instrument for studying the physical properties of polymers, and it may be used to identify the micro-morphology and surface of EPS. The surface structure of EPS-SL and EPS-N5 are shown in Fig. 5. The EPS-SL indicated a three-dimensional architecture with the film surface less compact, richer, and smoother alike a lotus leaf (Fig. 5A and B). L. Wang et al. (2021) conclude that these properties imply the EPS potential of industrial applications for food packaging films and plasticized films. Comparing, the EPS-N5 three-dimensional structure upon increased magnification, revealed a stiff-like, more spherical structure. The EPS feature porous characteristic (Fig. 5C and D). A similar porous structure was also reported by the EPS of Lactobacillus plantarum YW11 and Natronotalea sambharensis sp. nov. (Singh et al., 2019, Wang et al., 2015a). These previous reported studies concluded that because EPS has a more porous and concentrated distribution structure, it may increase its water retention capacity and retain more moisture, thereby expanding the physical properties of the product by generating a hydrated polymer with matrix-like properties and increasing viscosity.

Fig. 5

Scanning electron micro image of EPS-SL (A: 2500×, B: 1000×) and EPS-N5 (C: 2500×, D: 1000×).

Biological activity

The EPS-SL and EPS-N5 have bioactivities such as antibacterial and antioxidant activity. The MIC method was used for antibacterial activity and 4 different methods such as DPPH, ABTS, phosphomolybdenum assay and the antioxidant activity of pure EPS-SL and EPS-N5 was investigated using reducing power.

Antioxidant activity

At various doses, the DPPH scavenging activity of EPS-SL and EPS-N5 was tested (Fig. 6A). The scavenging activity of EPS-SL and EPS-N5 increased with increasing concentrations, according to the findings. At 5 mg/mL, EPS-SL and EPS-N5 had maximal DPPH radical scavenging activity of 52.7 and 64.9 %, respectively. Furthermore, the EPS-N5 had a higher scavenging activity than EPS-SL, and both were lower than vit C. At concentrations of 8 and 10 mg/mL, respectively, EPS generated by Bacillus velezensis SN −1 and Lactobacillus plantarum JLAU103 had the highest DPPH scavenging ability (Cao et al., 2020, Min et al., 2019). This antioxidant capacity of EPS can be attributed to the transfer of hydrogen by the polysaccharide to DPPH while interacting with its radical (Fooladi et al., 2019).

Fig. 6

Antioxidant activity of EPSs with different techniques (A: DPPH, B: ABTS, C: Phosphomolybdenum assay, D: Reducing power).

Another method for determining the antioxidant activity of natural items is to utilize the ABTS radical scavenging method. As shown in the Fig. 6B, the scavenging of ABTS radical multiplied with the increase of EPS-SL and EPS-N5 and the maximum clearance was 60.8 and 75.2%, respectively, when the concentration was 5 mg/mL. Ayyash et al observed a lower value than found by our study, with a clearance of 18.49% of EPS produced by Lactococcus gravieae C47 (Ayyash et al., 2020). Our results were close to those found by Bomfim et al and Xu et al (Bomfim et al., 2020, Xu et al., 2019).

The ABTS radical scavenging activities of EPS-SL and EPS-N5 samples outperformed those of DPPH free radical scavenging. This discrepancy might be explained by the fact that ABTS is a hydrophilic molecule, whereas DPPH is hydrophobic, according to the authors. The reducing power, as an important characteristic of antioxidant activity for natural product. In this assay, the antioxidant agent reduces ferricyanide (Fe3+) to the ferrous (Fe2+) form which was measured at 700 nm. The reducing power of the EPS-SL, EPS-N5 and vit C was shown in Fig. 6D. At 5 mg/mL EPS-SL and EPS-N5 showed a reducing power of 1.08 and 1.26, respectively. Like the other antioxidant test, the reducing power was correlated with its concentrations. In this study, the phosphomolbdenum assay was also studied. This approach is based on antioxidant agents reducing phosphate-molybdenum (VI) to phosphate-molybdenum (V) and forming a green complex at an acidic pH (Choudhuri et al., 2021). The reduction of Mo (VI) increased with the increase of concentration of EPS-SL and EPS-N5, and the maximum reduction was 0.594 and 1.0105, respectively (Fig. 6C).

Antibacterial activity

The antibacterial activity of EPS-SL and EPS-N5 was studied using the minimal inhibitory concentration (MIC) method. The results showed that EPS-SL and EPS-N5 had an interesting antibacterial activity. The MIC of both EPS was 1.75, 2.5 and 5 mg/mL against S. aureus, E. coli and L. monocytogenes, respectively. No effect was observed against S. pyogenes and S. thyphimurium. In comparison, Ye et al found a MIC of 2 and 3 mg/mL against E. coli and S. aureus (Ye et al., 2019). Fooladi et al observed no effect against L. monocytogenes and found a 3.3 mg/mL MIC against S. aureus (Fooladi et al., 2019). The mechanism of antibacterial action may be attributed to the interaction between the EPS and peptidoglycan and disrupt the structure, or the EPS interfere with peptidoglycan synthesis enzyme at substrate level, and stopping the cell wall biosynthesis and disruption. A few studies suggest that the EPS can block channels or receptors on the outer membrane (Abdalla et al., 2021). Another study, report that the EPS facilitate the accumulation of secondary metabolites, which could negatively affect pathogens (Salachna et al., 2018).

Conclusions

In this research, EPS producing strains, isolated from raw donkey milk, was reported as Leuconosctoc mesenteroides SL and Enterococcus vikkiensis N5. According to the RSM findings, the optimized medium containing sucrose, peptone, and inoculum produced the most EPS. The results of GC–MS, FTIR, and NMR analysis confirmed that the EPS-SL and EPS-N5 are heteropolysaccharides containing various monosaccharide components connected by α-(1 → 6) and -(1 → 3) linkages. The SEM image revealed that the EPS-SL showed a smooth and a lotus leaf shape, and EPS-N5 revealed a stiff-like, porous and more compact than EPS-SL. In addition, EPS-SL and EPS-N5 showed a higher degradation temperature, 280 °C and 340 °C, respectively. Moreover, EPS-SL and EPS-N5 showed an antibacterial activity against S. aureus, E. coli and L. monocytogenes with a MIC of 1.75, 2.5 and 5 mg/mL, respectively, and also showed an antioxidant activity. Our study suggests that EPS-SL and EPS-N5 may be a potential alternative as a food additive with an important antioxidant and antibacterial activity.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.