Biotechnological potential and genomic analysis of the Leuconostoc mesenteroides F17 and F18 strains isolated from camel milk

Abstract

Camel milk is known for its distinctive nutritional and antimicrobial properties. Thus, it offers a valuable niche for discovering novel probiotic strains with potential health benefits. In this context, two Leuconostoc mesenteroides strains, F17 and F18 were isolated from this dairy matrix and their properties analysed. Also, to gain deeper insight into their probiotic and technological traits, the sequencing of the whole genomes of both strains was conducted, revealing key genetic features supporting their suitability as starter cultures. The two bacteria displayed desirable technological properties including among them rapid milk acidification capability as well as proteolytic and lipolytic activities. F17 and F18 produced riboflavin (0.11 and 0.15 mg/L) and exopolysaccharides (1.8 g/L and 3.7 g/L) characterized as dextran by physicochemical analysis. The F17 and F18 DsrD dextransucrases of ~ 170 kDa, responsible for the dextran synthesis, were visualized by a zymogram analysis and their dsrD coding genes, with 98.99% identity, were identified in their genomes. After 24 h growth in a sucrose-containing medium, analysis of metabolic fluxes in culture supernatants of both bacteria demonstrated accumulation of the sweetener mannitol and the antimicrobial lactic acid; at the levels of 59 mM and 76.57 mM in the case of the best performer (F18). Furthermore, the results demonstrated that both strains exhibited good survival rates under in vitro simulated stomach duodenum-passage. A high surface adhesion was also observed for F17 and F18 including hydrophobicity of 30.15% and 31.72% and auto-aggregation ability of 44.71% and 53.23%, with a good resistance to phenol exposure and to moderated heat shock. As expected from the genomes inspection, the safety analysis revealed that the two strains were susceptible to most of the tested antibiotics, and none of them showed any hemolytic activity or biogenic amines synthesis capability. Both cultures supernatants exhibited significant growth inhibition of foodborne pathogens, presumably due to the present of lactic acid. Thus, these findings suggest that camel milk may harbor lactic acid bacteria with promising technological and probiotic traits, as exemplified by F17 and F18.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12866-025-04349-x.

Background

Foods are suitable environments for the growth of various microorganisms rendering them inedible and even dangerous for human or animal consumption. On the other hand, the fermentation of some food matrices has become crucial in food industry for their perceived health benefits and consumers liking [1].

Lactic acid bacteria (LAB) have been traditionally used to improve the safety and organoleptic characteristics of fermented foods. This group of bacteria is generally considered safe and scarce reports have shown their spoilage and pathogenic potential [2]. Thus, these bacteria are important for the food industry due also to their capability to produce several bioactive metabolites such as lactic acid, exopolysaccharides (EPS), ethanol, vitamins, hydrogen peroxide, diacetyl, mannitol, acetaldehyde and bacteriocins [3]. These LAB metabolites enhance the functionality as well as the safety and the organoleptic and rheological properties of the foods [2].

In addition, these bacteria can also improve the intestinal microbiota, protect against infections, alleviate lactose intolerance, reduce blood cholesterol levels, have bifidogenic effects and stimulate the immune system [2], and these benefits are not limited to live probiotic bacteria, since many studies demonstrate that even inactivated bacterial cells or their metabolites can also possess significant health-promoting properties [4].

In this context, among the EPS synthesized by LAB, the homopolysaccharide dextran has potential health effects for human beings such as antitumor, antioxidant, prebiotic, antimicrobial, and cholesterol-lowering activities as well as immunomodulatory potential [5]. In addition, this EPS acts as emulsification, thickening and gelling agent in food industry [6], and it could be used to develop functional food.

Concerning to other postbiotic compounds, some LAB could synthesize some vitamins within the B group. Among them, the riboflavin (vitamin B₂), which is essential for human beings, and it has to be acquired in the diet. Moreover, the vitamin B₂ deficiency is known as ariboflavinosis, and it can lead to health issues such as liver and skin damage, neurological changes, and impaired glucose metabolism [7]. Therefore, the capability to produce this vitamin could be another advisable probiotic property for a LAB.

The exploitation of LAB as starter cultures in food fermentation is intrinsically linked to the composition of the microbial consortium employed. Thus, although homofermentative LAB are the most common and widely used in food industry, recently, obligate heterofermentative LAB including Leuconostoc and Weissella, have gained interest as starter cultures and investigations have highlighted the potential of these microorganisms [8].

Camel milk represents a valuable food source for nomadic people in arid zones of Africa and Asia. In addition to its nutritional value, several studies have revealed the health-promoting benefits of camel milk. This is why the attention toward this type of milk is increasing. Its unique composition in bioactive components including immunoglobulins, lactoferrin, lysozyme and vitamins, have been associated with several physiological benefits after its ingestion for human beings such as antidiabetic, antihypertensive, antiallergic, carcinopreventive, antioxidant and immunomodulatory properties [9]. Nevertheless, the microbial diversity of camel milk has been poorly investigated compared to other types of milk. However, some studies have reported that camel milk has significant potential to provide probiotics LAB, such as strains belonging to Lactococcus lactis [10], Lacticaseibacillus casei, Lactiplantibacillus plantarum, Leuconostoc mesenteroides and Pediococcus pentosaceus species [11].

In this context, this study was mainly designed to investigate the probiotic and technological properties of two LAB strains isolated from Algerian raw camel milk and belonging to the L. mesenteroides species with the future aim of using them as starters or adjunct for the development of new functional food. The detection of these beneficial properties had encompassed a comprehensive investigation into their genomic features and fermentative behaviours.

Materials and methods

Bacteria isolated and used in this work and growth conditions

A total of 124 LAB isolates were recovered from three Algerian raw camel milk samples aseptically collected in February 2019 as well as in June and September of 2021 from three different regions in Algeria (Biskra, Ghardaia and Oran), using Man, Rogosa and Sharpe medium (MRS, BD Difco™, France) containing either 2% glucose (MRSG) or 2% sucrose (MRSS), and Mayeux, Sandine and Elliker (MSE) medium [12], supplemented with 10% sucrose and 30 µg/mL vancomycin [13], and after growth at 30 °C for 72 h. The isolates were purified by streaking on MRSG supplemented with 1.5% agar (MRSG-agar) and then stored at −80 °C in MRSG medium supplemented with 20% glycerol for further studies.

Only two isolates designated F17 and F18, with ropy phenotype in MRSS supplemented with 1.5% agar (MRSS-agar), were selected due to their ability to grow in a chemically defined medium (CDM) [14] lacking riboflavin. The metabolic fluxes as well as the quantification of the EPS and the riboflavin produced by the two isolates were investigated upon growth in Riboflavin Assay Medium (RAM, BD Difco™,France) lacking riboflavin and containing 2% glucose (RAMG) and in this medium supplemented with 2% sucrose (RAMGS). Brain Heart Infusion medium (BHI, Difco) supplemented with 1.5% agar (BHI-agar) and with (5%) defibrinated human blood was used to test a potential haemolytic activity. Mueller-Hinton medium supplemented with 1.5% agar (MH-agar, Difco) was used to test production of antimicrobial compounds.

The Escherichia coli V517 strain [15] carrying 8 plasmids (pVA517A through VA517H) was grown in Luria Bertani broth (LB, Sigma-Aldrich) and its extracted plasmids were used as plasmid size standard.

The foodborne pathogens indicator strains E. coli ATCC25922, Listeria ivanovii ATCC 19,119, Pseudomonas aeruginosa ATCC 27,853, and Staphylococcus aureus ATCC 25,923 were used in this study to evaluate the antimicrobial potential of the two isolates in MH-agar, after sub-culturing in BHI at 37 °C for 18 h.

Phenotypic and physiological characterization

In order to identify the two isolates at the genus level, only Gram-positive and catalase negative isolates with ropy phenotype in MRSS-agar were subjected to diverse biochemical and morphological tests as previously described [16], including: gas production from glucose, arginine hydrolysis, citrate metabolism, acetoin production as well as growth at different temperature (4 °C, 15 °C, 37 °C and 45 °C), or in the presence of two NaCl concentrations (3.0% and 6.5%) or at different pH values (4.0 and 8.0). Furthermore, to identify the isolates at the species level, their carbohydrates fermentation patterns were analysed using API 50 CHL test kit (BioMérieux, France).

The cell morphology and detection of the EPS production was investigated by transmission electron microscopy (TEM) using a JEOL JEM-1230 electron microscope (JEOL Ltd., Tokyo, Japan) operating at an accelerating voltage of 100 kV.

Genotypic characterization

Genomic DNA extraction, sequencing, assembly, annotation and analysis

The L. mesenteroides F17 and F18 strains were grown until early exponential phase to an optical density at 600 nm wavelength (OD_{600 nm}) of 0.6. Then, the genomic DNA of both strains was extracted using the Wizard Genomic DNA Purification kit (Promega, USA) following the manufacturer’s instructions with few modifications, as previously described [17]. Afterwards, the DNA was purified with NucleoSpin Gel and PCR Clean-up Kit (Macherey Nagel, Düren, Germany) according to the manufacturer’s instruction. The DNA concentration was determined by measurement of the fluorescence using the QubitTM dsDNA BR Assay Kit (Invitrogen, USA) and the Qubit^® 2.0 fluorometric detection methods (Thermo Fisher Scientific, Waltham, MA, USA). Then, the integrity of the genomic DNA was visualized by staining with GelRed 1X (Biotium, USA) upon fractionation in 0.8% agarose gel.

Sequencing of the entire genome was performed at Secugen (Madrid, Spain) using NGS Illumina Miseq technology with 2 × 150 paired-end reads in a Ilumina NextSeq combined with NGS Oxford Nanopore MinION technology. Libraries were prepared with the SQK-LSK109 ligation kit (Nanopore Technologies, UK). A label was added to each sample (barcode) with the native barcoding kit EXP-NBD114 and the libraries were loaded in the flow cell FLO-MIN106 of a MinION equipment (Nanopore Technologies). The Illumina and Nanopore reads were analysed with a high quality module (super accurate) of the MinKNOW software. The assembly was performed with Galaxy unicycler 0.5.0. software [18]. The DNA sequence of the chromosomes and the plasmids present in the F17 and F18 strains have being deposited in the NCBI GenBank and their deposit references are depicted in Table 1.

Table 1.

Accesion numbers of the genomes of L. mesenteroides F17 and F18 strains in the GenBank

SUBID	BioProject	BioSample	Location	Accesion nº	Organism
SUB15029791	PRJNA924563	SAMN46388514	Chromosome	CP178852	L. mesenteroides F17
SUB15029791	PRJNA924563	SAMN46388514	pLC1	CP178853	L. mesenteroides F17
SUB15029791	PRJNA924563	SAMN46388514	pLC2	CP178854	L. mesenteroides F17
SUB15029791	PRJNA924563	SAMN46388514	pLC4	CP178855	L. mesenteroides F17
SUB15029791	PRJNA924563	SAMN46388514	pLC6	CP178856	L. mesenteroides F17
SUB15032603	PRJNA924563	SAMN46394793	Chromosome	CP178911	L. mesenteroides F18
SUB15032603	PRJNA924563	SAMN46394793	pLC3	CP178912	L. mesenteroides F18
SUB15032603	PRJNA924563	SAMN46394793	pLC5	CP178913	L. mesenteroides F18

The genomes annotations were performed using Rapid Annotation and Subsystem Technology (RAST) [19], the Prokaryotic Genome Annotation Pipeline (PGAP) at NCBI [20], and Prokka 1.14.6 tool at Galaxy web server [21]. Proksee bioinformatic tool were used for genome mapping visualization [22]. The putative origin of replication (Ori) of the plasmids was predicted using the Ori-Finder 2022 platform [23], with parameters adjusted to specific DNA boxes of Gram positive bacteria (Bacillus subtilis). ResFinder 4.6.0 database and the gene homology were confirmed by Blast tools. The putative open reading frames (ORF) encoding-proteins were predicted using (i) SeqBuilder Pro (DNASTAR, Lasergene 17, Madison, WI, USA), (ii) ORFfinder at NCBI and (iii) pLannotate webserver. The nucleotide and protein-coding sequences homology were determined by Blastp tool at NCBI. The theorical molecular weight (Mw) of proteins was predicted using Expasy Compute Mw tool. The prediction of antimicrobial resistance phenotypes, genomic islands and virulence factors were performed using Resistance Gene Identifier (RGI) tool from the Comprehensive Antibiotic Resistance Database CARD web server [24].

Plasmid profiling

Total plasmidic DNA preparations of L. mesenteroides F17 and F18 strains and of E. coli V517 were obtained as previously described [25] with some modifications for the LAB. Briefly, 5 mL of exponential bacterial cultures grown in MRS to an OD_{600 nm} of 1.0, approximately 10⁸ colony forming units (CFU/mL) were sedimented by centrifugation (at 8,000 × g for 10 min, 4 °C), resuspended in 200 µL of solution I (25% sucrose and 30 mg/mL lysozyme) and lysed by incubation for 15 min at 37 °C. Then, 400 µL of solution II containing 3% sodium dodecyl sulphate (SDS) and 0.2 N NaOH were added and samples were incubated for 7 min at room temperature (rt) to disrupt the cytoplasmic membrane and denature the chromosomal DNA. Furthermore, upon addition of 3 M sodium acetate pH 4.8 (300 µL) and mixing, the samples were centrifugated (at 10,000 × g, 15 min, 4 °C). Afterwards, DNA precipitation from the supernatants and further concentration was performed by addition of isopropanol (650 µL) and centrifugation as above. The sedimented DNA was resuspended in distilled water (320 µL), then deproteinated by addition of 200 µl of solution III (7.5 M ammonium acetate, ethidium bromide (0.5 mg/mL) plus 350 µl of (1:1) phenol: (24) chloroform-(1) isoamylic alcohol and mixing, followed by recovering by centrifugation (at 10,000 × g, 5 min, rt). Afterwards, the aqueous phase was mixed with ethanol (1 mL) and the DNA was further precipitated and recovered by storage overnight at −20 °C and centrifugation (at 10,000 × g, 15 min, 4 °C). The sedimented plasmidic DNA was washed with 70% ethanol re-centrifuged, and finally resuspended in 10 mM Tris HCl pH 8.0. Then, the plasmids were visualized by staining with GelRed upon fractionation in 0.8% agarose gel. To estimate the plasmidic size and conformation, it was also run in the gel an E. coli V517 plasmidic preparation carrying 8 plasmids (pVA517A through pVA517H). Gel Doc 2000 Bio-Rad gel documentation system (West Berkeley, USA) and the Quantity One 4.5.2 Bio-Rad software were used to capture and analyse the images.

Detection and quantification of riboflavin

Cultures of the F17 and F18 strains were grown overnight, then the bacterial cells were sedimented by centrifugation (at 10,000 × g, 10 min, 4 °C) and washed twice with saline solution under the same conditions. Afterwards, each bacterial pellet was used to inoculate both RAMGS and RAMG media and further grown at 30 °C. When the cultures reached an OD_{600 nm} of 1.0, they were diluted (at a 1:10 ratio) in the corresponding fresh medium, 200 µL aliquots of the diluted cultures were dispensed in triplicate in 96-well polystyrene optical bottom plate (Thermo Fisher Scientific, Rochester, NY, USA) Then, the OD_{600 nm} and the riboflavin fluorescence were monitored in real time every 30 min, at 30 °C until the cultures reached the late stationary phase using a Varioskan Flask System (Thermo Fisher Scientific, Waltham, MA, USA). The riboflavin fluorescence was measured upon excitation at a wave length of 440 nm and detection of emission at a wavelength of 520 nm, as previously described [26]. The riboflavin concentration was calculated using a calibration curve of riboflavin dissolved in RAMG and measuring the fluorescence of the riboflavin solutions. All the experiments were performed in triplicate.

Analysis of the EPS synthesized by F17 and F18 strains

Production, purification and quantification of EPS

The two LAB strains were grown in MRSG and MRSGS to an OD_{600 nm} of 1.0, recovered and washed by centrifugation as indicated in section “Detection and quantification of riboflavin”. Then, the bacterial cells were resuspended in RAMG or RAMGS, respectively and further incubated at 30 °C until OD_{600 nm} of 1.0. The RAMGS cultures were used to inoculate 40 mL of fresh RAMGS with an initial OD_{600 nm} of 0.01 and incubated for 16 h at 30 °C. Afterwards, the EPS concentration was determined as previously described [31] by the phenol sulphuric method [27] using a glucose calibration curve.

To purify the polymers produced by F17 and F18, they were isolated from the supernatants as previously described [28]. Briefly, the cultures supernatants were recovered by centrifugation (at 13,000 x g, 1 h, 4 °C), then mixed with cold absolute ethanol (v/v) and maintained at 4 °C for 24 h. The precipitated EPS were sedimented by centrifugation as above, dried, resuspended in ultrapure water and dialyzed against water using a membrane with a 12–14 kDa cut-off for 72 h at rt. Finally, the EPS were lyophilized and kept dry at rt until further characterization.

To stablish the EPS concentration and purity, samples obtained in each step were used to measure the total neutral sugar content by the phenol–sulfuric method, and the concentrations of the potential contaminants DNA, RNA and proteins by specific fluorescent staining kits and the Qubit^® 2.0 fluorometric detection methods.

Characterization of the produced EPS

After purification, the polymers produced by F17 and F18 strains were characterized as previously described [29] determining: (i) their monosaccharide composition after acid hydrolysis by gas-chromatography mass spectroscopy (GC-MS), (ii) their anomeric configuration by Fourier transform infrared (FT-IR) spectroscopy and (iii) their linkage types by methylation analysis.

Detection of dextransucrases (Dsr) activities

Dsr activity was investigated by in situ polymer production on gel [30, 31]. Briefly, F17 and the F18 strains were grown at 30 °C for 16 h in RAMG and RAMGS. Then, the culture supernatants were recovered after centrifugation (at 10,000 × g, 10 min, 4 °C), loaded onto a SDS-polyacrylamide gradient gel (5–8%) and fractionated by electrophoresis at 21 °C with constant voltage (100 V). To reveal Dsr activities, after three-fold washing with 20 mM sodium acetate buffer (pH 5.4) supplemented with CaCl₂ (0.05 g/L) and 0.1% Triton X-100, in situ dextran synthesis was induced by 16 h incubation at 21 °C in sodium acetate buffer supplemented with 10% sucrose. Afterwards, the gel was subjected to two incubations: (i) in 75% ethanol for 30 min and (ii) in a solution containing 0.7% periodic acid and 5% acetic acid for 1 h. Next, the gel was washed three times for 20 min using a solution containing 0.2% sodium metabisulfite and 5% acetic acid. Finally, the Dsr activity was revealed by exposing the gel to the periodic acid-Schiff’s reagent. The Mw marker Pre-stained Precision Plus Protein (Bio-Rad) including polypeptides in the range of 10–250 kDa was also loaded in the gel to estimate the Mw of the active Dsr enzymes detected.

Analysis of sucrose metabolism

The metabolic behaviour of the two LAB in the presence of sucrose was investigated through metabolic fluxes analysis as previously described [32]. Briefly, F17 and F18 strains were grown in MRSS until OD_{600 nm} of 1.0, sedimented and washed as indicated in section “Detection and quantification of riboflavin”. Then, they were resuspended in fresh RAMGS and incubated at 30 °C until OD_{600 nm} of 1.0. At this point the cultures were diluted 1:100 in fresh RAMGS and grown for 24 h at 30 °C. During the first 7 h and at the end of the incubation period (24 h), there were monitored: (i) the evolution of the pH using a pH meter (Crison, Spain) and (ii) the cell growth by determination of the CFU/mL by plating of the cultures on MRSG-agar. Also, the recovered supernatants, upon centrifugation of the cultures as described above, were used to determine: (i) the EPS concentration by the phenol-sulphuric acid method, and (ii) the concentration of glucose, sucrose fructose, mannitol and lactic acid by GC-MS using myo-inositol as internal standard. All experiments were performed in triplicate.

In vitro evaluation of technological and probiotic characteristics

Acidification and coagulation ability

The acidification of skim milk by F17 and F18 strains and the curd formation of the drink after bacterial fermentation were tested as previously described [33], with slight modifications. Briefly, 10% reconstituted skim milk was enriched with 0.3% (v/v) yeast extract and 2% glucose. The acidification ability was evaluated in kinetic curves during incubation at 30º C for 72 h by measurements of pH every 2 h. The ability of strains to coagulate milk was revealed by the appearance of curd and cracks.

Screening for enzymatic activities

The F17 and F18 strains were examined for their lipolytic activity, using a buffered MRSG-agar at pH 7.0, opacified with 0.5% (w/v) CaCO₃ in order to visualize a possible lipase activity [16]. Then, the medium was supplemented with 1%, 3% or 5% (v/v) of artificial or natural lipid source (olive oil and tween 20) as substrates. After 48 h of incubation at 30 °C, a clear zone surrounding the inoculated spots indicates a lipolytic activity.

The proteolytic activity test was carried out as described in [16], by spotting 5 µL of each exponential bacterial culture on PCA-agar medium supplemented with 1% or 5% reconstituted skim milk. After incubation at 30 °C for 24 h and 48 h, the proteolytic activity was expressed as the diameter of the clear zone around the inoculated spots (mm), that arises as a result of hydrolysis of the caseins.

Resistance to gastrointestinal (GIT) stresses

Approximately 10⁸ CFU/mL of either F17 or F18 strains from a culture grown in MRSG medium to an OD_{600 nm} of 1.0 were harvested by centrifugation (at 8,000 × g, 10 min, 4 °C), washed, and inoculated in different simulated gastric and intestinal matrices as previously described [16, 34] exposing the bacteria to: (i) pH 3.0 in the presence of the digestive enzyme pepsine (3 mg/mL), (ii) bile salts at 0.3%, 1.0%, or 2.0%. Samples of the treated bacteria were taken at time 0 h (t₀) and after 3 h (t₃). Then, the CFU were determined by the platting method using MRSG-agar, and incubation at 30 °C for 48 h.

The experiment was performed in triplicate and the cells survival after 3 h incubation (t₃) was expressed as percentage considering as 100% the CFU detected at (t₀) according to the following Eq:

1

Response to simulated stomach duodenum-passage

The behaviour of F17 and F18 strains in an in vitro model was evaluated as previously described [16], Briefly, the pelleted and washed bacterial cells as detailed above were diluted (at a 1:10 ratio) using a Ringer’s solution containing sodium chloride (9 g/L), potassium chloride (0.42 g/L), calcium chloride (0.48 g/L) and sodium bicarbonate (0.2 g/L), and used to inoculate (at a 1:3 ratio) MRSG medium at pH 3.0. Experiments were carried out in triplicate and the percentage of cells survival was determined as indicated in section “Resistance to GIT stresses”.

Bacterial hydrophobicity

The sedimented cells from the F17 and F18 cultures, grown as indicated above, were resuspended in 3 mL of phosphate buffered saline (PBS) pH 6.5, adjusted to an initial OD_{600 nm} of 1.0, and mixed with 0.6 mL of hydrocarbon solvents (xylene). Then, the mixture was incubated at 37 °C for 10 min and stirred with vortex for 2 min. After maintaining the mixture at rt for 15 min without agitation, the absorbance value of the aqueous phase was measured at a wavelength of 600 nm. The protocol described was carried out with three independent replicates. The cell surface hydrophobicity was evaluated in percent using the following formula [16]:

2

Auto-aggregation

The auto-aggregation ability of F17 and F18 strains was investigated using a method previously described [34]. Briefly, sedimented bacterial cells obtained as described above, were suspended in 5 mL of PBS (approximately 10⁸ CFU/mL) and stirred by vortexing for 10 s. Then, the bacterial suspensions were incubated for 3 h at 30 °C. Afterwards, 0.1 mL, from the top of the suspensions containing the non-aggregated cells were gently pipetted, and their absorbance at a wavelength of 600 nm was measured (At₃). Also the absorbances of the suspensions at time 0 h (At₀) were measured. The auto-aggregation coefficient was calculated using the following equation:

3

Experiments were performed in triplicate.

Production of antimicrobial compounds

The antimicrobial activity of F17 and F18 strains was determined using the well diffusion test according to [34]. In brief, fresh cultures grown at OD_{600 nm} of 1.0 were subjected to centrifugation (at 10,000 × g, 10 min, 4 °C), and 100 µl of the recovered supernatants were introduced into wells previously performed in MH-agar plates pre-seeded with each foodborne pathogen indicator strain at 1 × 10⁷ CFU/mL, then the plates were incubated at 37 °C for 24 h. The inhibitory zone diameter around each well was measured and expressed in mm. In addition, three other aliquots of the same supernatants were treated before being introduced into the wells. The first sample was neutralized using 1 N HCl, while the other two samples were subjected, after neutralization, to enzymatic treatments using 1 mg/mL of proteinase K or pepsin. After the enzymes addition, the samples were incubated at 37 °C for 1 h, and subsequently sterilized by passage through a 0.22 μm Millipore filter. Then, 100 µl of the treated samples were spoted on MH-agar pre-seeded with pathogens as indicated above and incubated at 37 °C for 24 h.

Test for safety considerations of the strains

Bacterial sensitivity to antibiotics

The antibiotic susceptibility of the F17 and F18 strains was assessed using the kirby-bauer disc diffusion assay according to the guidelines of the Spanish Antibiogram Committee [35], applying the widely used commercially-available antibiotics. Briefly, fresh cultures grown to an OD_{600 nm} of 1.0 were diluted (at a 1:10 ratio) to correspond with 0.5 McFarland. Then, they were seeded on MH-agar plates, and the antibiotic-impregnated disks were applied on the agar surface. Antibiotic susceptibility was assessed by measuring the inhibition zone diameters around the discs (mm) after 24 h incubation at 30 °C. Results were interpreted as sensitive (S), intermediate (I), or resistant (R) to the antimicrobial agent, according to the interpretative zone antimicrobial diameters standards as previously reported [36, 37].

Production of biogenic amines (BA)

The production of BA was tested by spotting 5 µL of each exponential bacterial culture at OD_{600 nm} of 0.6 on the Bover-Cid agar medium [38], that contains pyridoxal-5-phosphate (0.005%), which is the cofactor for amino acid decarboxylase activity, plus 0.06% bromocresol purple as pH indicator and 0.5% of each amino acid precursors L-tyrosine, L-ornithine or L-histidine responsible for synthesis of tyramine, putrescine or histamine, respectively. The decarboxylation of amino acids appears as a purple colour around the spot which indicates the production of BA.

Hemolytic activity

The hemolytic activity was assessed on BHI-agar supplemented with defibrinated human blood. The hemolytic reaction was optically evaluated after growth for 48 h at 30 °C, and distinguished as α-hemolysis (green zone), β-hemolysis (clear zone), or γ-hemolysis (no reaction) [39].

Statistical analysis

The probiotic features analyses were performed at three independent experiments and results presented as mean values and corresponding standard deviation. The data were subjected to one-way analysis of variance (ANOVA) using the the R software version 4.3.0 [40]. Results are presented with letters and significant difference were determined at p < 0.05. Means with the same letter are not significantly different.

Results and discussion

Selection and phenotypical identification of EPS- and riboflavin-producing L. mesenteroides strains

In this work, we have approached the identification of EPS- and riboflavin-producing LAB isolated from camel milk. To this aim, a total of 124 isolates from three Algerian raw camel milk samples were analysed here. All the isolates exhibited a typical colony morphology of LAB, did not produce catalase and were Gram-positive (results not shown).

Among them, two isolates designated F17 and F18 were selected as potential EPS producers in presence of sucrose, because they generate colonies with a ropy or mucoid phenotype (Figs. 1 A and 1 C) in MRSS-agar medium, that was not detected upon growth on MRSG-agar medium containing glucose instead of sucrose (Fig. 1B and D).

Fig. 1.

Macroscopic detection of EPS production by F17 and F18 strains. The bacteria were grown on MRSS-agar (A) and (C) or on MRSG-agar (B) and (D) at the indicated times and temperatures. The depicted photographs show production of EPS by both bacteria only on MRSS-agar

In addition, the ropy phenotype was observed for the two bacteria in a wide range of temperatures from 15 °C to 30 °C, being the mucosity of the colonies more pronounced for F18 (Fig. 1 C vs. 1 A).

Furthermore, we have tested whether the two isolates were prototrophic for riboflavin due to their capability to produce this vitamin. The bacteria were inoculated in RAMGS and RAMG liquid media lacking the vitamin B₂, and their growth performance in real time was monitored by measuring the increase of the OD_{600 nm} (Fig. 2). Also, the production of riboflavin in the bacterial cultures was detected during growth by measuring the fluorescence of the vitamin (Fig. 2).

Fig. 2.

Detection of riboflavin production by F17 an F18 strains. Bacteria were grown in RAMGS (GS) or RAMG (G) at 30 °C, and the growth of the cultures (OD_{600 nm}) as well as the production of riboflavin (its fluorescence) were measured in real time during 15.5 h. The represented values are the mean of three independent experiments. In both media, efficient growth accompained by riboflavin production was observed for the two bacteria

Both bacteria were able to grow in both media and they produced riboflavin. The two isolates reached a higher final OD_{600 nm} in medium containing sucrose in addition to glucose (Figs. 2A and 2C vs. Figure 2B and D, respectively). These results support that in the two bacteria the sucrose utilization improves the growth due to the glucose metabolism. Production of riboflavin took place in both media, but it was more efficient in RAMGS (Figs. 2 A and 2 B vs. Figures 2 C and 2 D, respectively), and it was higher for F18 (Fig. 2B and D) than for F17 (Figs. 2A and 2 C). Moreover, when riboflavin production was quantified after 15 h of growth (data from Fig. 2), it was detected that the two LAB reached final vitamin levels of: 0.15 ± 0.01 mg/L and 0.11 ± 0.00 mg/L or 0.11 ± 0.01 mg/L and 0.09 ± 0.01 mg/L in RAMGS and RAMG for F18 or F17, respectively. Thus, F18 produced slightly higher concentration of riboflavin than F17. These levels were lower than that previously detected for L. mesenteroides and Liquorilactobacillus mali strains (~ 0.2 mg/L) isolated from Algerian food products [31] and for Weissella confusa FS54 (~ 0.25 mg/L) [41] wild-type strains grown in the same media. Moreover, the production was substantialy lower than that of strains selected by treatment with roseoflavin, i.e. W. confusa FS45 B2 overproducer of 4.9 mg/L of riboflavin [41]).

It was also observed that production of riboflavin by F17 and F18 in RAMGS (Figs. 2 A and 2B), and at less extend in RAMG (Figs. 2 C and 2D), did not take place at the beginning of the exponential phase, although both bacteria started to growth from the beginning of the incubation. This behaviour has also been observed in previous studies for wild-type strains belonging to the Leuconostoc and Weissella genera [17, 31, 41], and it could be interpreted as a need of consumption of the intracellular riboflavin prior relieve of repression of the ribDGAH (rib) operon expression encoding the enzymes required for the vitamin biosynthesis.

Concerning to the physiological and biochemical traits of F17 and F18 (Supplementary Table S1), these strains have the ability to produce CO₂ from glucose, characteristic that closely related them to the heterolactic fermentation pathway. They were not able to catabolize arginine through the arginine dihydrolase pathway, and to use citrate in the presence of glucose. According to these results these two strains were initially identified at the genus level as Leuconostoc sp. In liquid medium the two bacteria grew well at 15 °C, 30 °C and 37 °C, however any of them was able to grow at 45 °C, and both grew at 4 °C after 5 days of incubation. In addition, the selected strains showed a high stability over the different range of pH values tested (from 4.8 to 9.0) and tolerate exposure to osmotic stress provided by NaCl at 3.0% and at 6.5% in MRSG medium. This good resistance to osmotic stress may be attributed to the high levels of salt and mineral content in camel milk compared to the bovine one [42]. Moreover, the carbohydrate assimilation profile of the two isolates was analysed using the API50 CHL kit and the results are depicted in Supplementary Table 2. The obtained pattern, together with the results reported above and according to the Bergey’s Manual of Systematic Bacteriology: The Firmicutes [43] revealed that F17 and F18 behave like the L. mesenteroides strains.

Sequencing and analysis of the entire genome of L. mesenteroides F17 and F18

The DNA sequences of the entire genomes of the F17 and F18 strains, including their chromosomes and plasmids, has been determined using NGS Illumina Miseq technology combined with NGS Oxford Nanopore MiniION technology, and deposited in the GenBank database under the accession numbers depicted in Table 1.

Chromosomal genome analysis

The whole chromosomal genome analysis including G + C content, and the number of coding sequences (CDS) is summarized in Fig. 3, and revealed that F18 has a genome with a length of 1,982,395 bp, slightly smaller than that of the F17 (2,046,746 bp). The vanT and vanY genes present in a vanG cluster for vancomycin resistance were detected with CARD analysis in the chromosomes of the two strains, in agreement with the previous knowledge of the intrinsic resistance of Leuconostoc spp. to this antibiotic [44].

Fig. 3.

Circular map representation of the F17 and F18 genomes. The maps were generated using proksee tool. Innermost rings show GC content and GC skew, the outer rings provide information about: forward coding sequences (CDS), reverse CDS, rRNA, tRNA, and tmRNA. These elements are indicated in different colours

Blast analysis of the DNA sequence of the F17 and F18 rrs genes encoding their 16 S RNA confirmed the phenotypical and biochemical typing, because the two sequences showed the highest similarity with those of the L. mesenteroides strains deposited in the data banks. In addition, multi-align comparison of the rrs genes of F17 and F18 with those of bacteria deposited in the GenBank allowed a construction of a phylogenetic tree (Fig. 4A and Supplementary Fig. S1A), which revealed that the F17 strain carries a rrs gene identical to those of L. mesenteroides subsp. dextranicum DSM20484 from Korean cheese and the L. mesenteroides subsp. mesenteroides CBA3650 isolated from cabbage kimchi inoculated with MRS-enriched fecal LAB. In addition, the rrs gene of F18 shears a 99.94% homology with them, and it is part of a distinct cluster with L. mesenteroides subsp. jonggajibkimchii DRC1506 isolated from Korean kimchi, as their rrs genes were identical. Also, in this cluster was present the rrs gene of L. mesenteroides subsp. suioanicum DSM 20,241, with a 99.81% homology This last bacteria has being reclassified as a new specie of the genus Leuconostoc based on genome comparisons [45]. The other two aligned genes, from strains belonging to Leuconostoc lactis and Leuconostoc citreum species (isolated from South Korean kimchi), were included in different sub-classes sharing lower homology (98.32–97.94%) with the 16 S rRNA coding genes of the Algerian L. mesenteroides strains (Fig. 4A). Furthermore, analysis of the chromosomal genome of the two strains revealed the presence of the rib operons, involved in riboflavin biosynthesis. Thus, these results also support that F17 and F18 are riboflavin producers. Also, multi-align comparison of the genes of the rib operons from F17 and F18 with those of bacteria deposited in the GenBank database allowed a construction of a phylogenetic tree (Fig. 4B and Supplementary Fig. S1B), which revealed that the genes of F17 and F18 have a 99.5% homology between them. Furthermore, F18 carries a rib operon identical to that of DRC1506, and the homolog operon of F17 has the highest similarity (95.53%) with that of the CBA3650 strain. Concerning to operons belonging to other species of Leuconostoc (kimchi, citreum, lactis and garlicum), they showed homology ranging from 68.33 to 62.6% with those of the Algerian strains. Finally, the rib operon of W. cibaria BAL3C-5 isolated from fermented Spanish dough, and whose riboflavin production has been investigated [17, 46], showed a homology of 59.51% and 59.41% with the genes of F17 and F18. Concerning to other vitamins of the B group, the folate (B₉ vitamin) and biotin (B₇) biosynthetic operons were also detected with RAST tool.

Fig. 4.

Phylogenetic trees of the rrs genes (A), rib operons (B) and dsrD genes (C) from F17 and F18 as well as close homologs from the GenBank database. The designation of the strains as well as the number of the genomes of the analyzed bacteria in the GenBank are indicated

Plasmid profiling

Concerning to the extrachromosomal elements, the genomes analysis revealed that F17 carries four plasmids named pLC1, pLC2, pLC4, pLC5, respectively, with sizes of 38,318 bp, 37,376 bp, 16,603 bp and 9,288 pb (Fig. 5), whereas F18 only harbours two plasmids of 32,262 bp and 4,707 bp designated pLC3 and pLC6, respectively (Fig. 5). In addition, analysis of plasmidic preparations of the two L. mesenteroides strains in an agarose gel (Fig. 6) revealed that indeed both carry the plasmids expected from the genome analysis (Fig. 5), showing that the assembling of the plasmids was correct.

Fig. 5.

Map prediction of the F17 and F18 plasmids. The maps were generated with SeqBuilder Pro. The genes homology was confirmed with Blastp tool and the putative origin of replication of the plasmids was predicted using the Ori-Finder 2022 platform with parameters adjusted to specific DNA boxes of Gram positive bacteria (Bacillus subtilis). Symbols: black box, putative replication origin Ori; arrows, putative open reading frames (ORF) encoding-proteins in both directions (light grey, forward; dark grey, reverse). Some predicted genes encoding proteins with less than 402 pb are not represented

Fig. 6.

Detection of plasmids carried by F17 and F18. Plasmidic DNA preparations of the L. mesenteroides strains were fractionated in a 0.8% agarose gel and plasmidic bands were revealed by straining with GelRed. As molecular weight standard (Mw) was used a plasmidic preparation of the E. coli V517 strain, among of its eight plasmids, seven were detected (pVA517A-pVA517G) and their known sizes are indicated in the left side of the figure.The analysis also revealed the covalently closed (CCC) as well as open circle (OC) or dimer forms of the six plasmids (pLC1-pLC6) carried by either F17 or F18, and their positions are indicated in the right side of the figure

Also, using SeqBuilder were identified the features of the plasmids and their maps are presented in Fig. 5. The plasmid profiling revealed that any of the five plasmids carry genes encoding proteins involved in antibiotic resistance and pathogenicity.

Furthermore, some of the genes present in the plasmids seem to encode proteins that should contribute to the capability of the strains to counteract environmental stresses, supporting also its potential utilization as probiotics. In brief, the pLC3 of F18 and the pLC2 of F17 carry genes encoding proteins involved in resistance to various stress conditions: heavy metal resistance (P-type ATPase), camphor resistance (CrcB protein) and DNA starvation/stationary phase protection (Dps protein).

Moreover, it was detected the copper resistance (multicopper oxidase) coding gene in the pLC5 of F17. Also, pLC4 plasmid should confer to F17 resistance to phage infection, since carries the genes encoding a type I restriction modification system as well as the abi gene involved in phage abortive infection like the gene present in pLC6 of F18, whose product should be responsible for phage resistance (protein belonging to the AIPR family).

Finally, the pLC4 and pLC5 plasmids from F18 carry genes encoding for conjugative relaxases belonging to the MOBv1 family, and involved in the process of DNA mobilization associated to bacterial conjugation.

Thus, the plasmids present in both bacteria confer to them beneficial properties for their use as starter in food industry.

Analysis and characterization of the EPS produced by the strains

The final aim of this part of the work was to perform the physicochemical characterization of the EPS produced by F17 and F18.

As a first step, the production of EPS in batch cultures was approached. We have previously shown that L. mesenteroides strains are able to produce EPS in the defined RAM medium suitable for their further purification without contamination of other polymers [31]. Therefore, the strains were grown in RAMGS and RAMG for 16 h and the production of EPS was quantified (Table 2).

Table 2.

Analysis of production of EPS by L: mesenteroides F17 and F18 strains

Strain	Medium	OD600 nm	Total EPS (g/L)	Ratio (EPS/OD)
L. mesenteroides F17	RAMG	2.0 ± 0.1c	ND
L. mesenteroides F17	RAMGS	2.5 ± 0.1b	1.8 ± 0.1b	0.7
L. mesenteroides F18	RAMG	1.9 ± 0.1c	ND	-
L. mesenteroides F18	RAMGS	3.4 ± 0.2a	3.7 ± 0.1a	1.1

ND, EPS was not detected

The experiments were performed in triplicate. The different letters indicate statistically significant differences with a p < 0.05 determined with a one-way ANOVA analysis

As expected from the results presented in Fig. 2, both strains grew better in RAMGS than in RAMG. Also, correlating with the phenotypical analysis (Fig. 1) both strains only produced EPS in the medium containing sucrose. Furthermore, analysis of bacterial cells of F17 and F18 cultures by TEM revealed the presence of EPS surrounding the bacterial cells only when sucrose was present in the growth medium (see below in Fig. 8A). In addition, in RAMGS, F18 produced statistically significant higher levels of EPS than F17 (3.7 g/L vs. 1.8 g/L). This difference was not due to the growth of the cultures, since a higher ratio of EPS production vs. OD_{600 nm} was also observed for F18 (1.0 vs. 0.7).

Fig. 8.

Detection of dextran production and DsrD activity of F17 and F18. (A) TEM pictures of F17 and F18 strains grown in RAMG and RAMGS are depicted. (B) A zymogram of culture supernatants of the L. mesenteroides strains grown in RAMG and RAMGS is shown. The analysis revealed that for the two LAB, upon growth in RAMGS: the bacterial cells are surrounded by EPS (A) and Dsr are present in cultures supernatants (B)

As a second step, the EPS produced by the two LAB was purified from the supernatants of the cultures grown in RAMGS by ethanol precipitation and dialysis to remove low Mw compounds present in the supernatant. Moreover, the yield of the EPS as well as the levels of contaminants (DNA, RNA and proteins) during the purification steps was determined (Supplementary Table S3). After purification, a final yield of 1.8 g/L and 2.3 g/L for the polymers of F17 and F18 with more than 98% purity was observed. In addition, only low concentrations of DNA, RNA and proteins were detected even in culture supernatants, and after the purification the EPS of both strains reached a purity of > 98% with respect to the other macromolecules analysed.

Finally, the polymers were subjected to physicochemical characterization (Fig. 7). Determination of their neutral sugar composition revealed that both EPS were composed of only glucose (Fig. 7B). Their IF-TIR spectrum showed that they were α-glucans, because they have an anomeric configuration type α corresponding to a register between 849 et 918 cm⁻¹ (Fig. 7A). Finally, methylation analysis (Fig. 7C) allowed to stablish that both polymers are dextrans with a primary structure (Fig. 7D) composed of (1,6)-D-glucopyranose units (77%−88%) in the main backbone and partially branched at positions O−4, O−2 and/or O−3, with D-glucopyranose units in the side chain. This main bone structure is characteristic of other dextrans [47] and this type with more than one branching has been previously observed in the polymers from other Algerian LAB isolated from food [31, 48].

Fig. 7.

Physicochemical analysis of the EPS synthesized by F17 and F18. Their anomeric configuration (A), their monosaccharide composition (B), and their linkage types (C) as well as their primary structure inferred (D) are depicted. The analysis revealed that the polymers of both LAB were dextrans with low percentage of ramifications

Detection in situ and in silico of Dsr

Dextran is synthesised in LAB in a reaction catalysed by extracellular Dsr using sucrose as substrate. The Dsr belong to the family of glycosyl hydrolyses 70, which upon hydrolysis of sucrose generate free fructose and dextran by elongation of glucopyranosyl chains. Moreover, zymogram analysis of supernatants of Leuconostoc cultures has allowed us previous detection of Dsr [30, 31]. Therefore, supernatants of cultures of F17 and F18 grown in RAMGS and RAMG were used to detect their Dsr enzyme(s) after fractionation in a SDS-polyacrylamide gel, removal of the SDS and incubation in presence of sucrose. The obtained zymogram profiles showed a similar band pattern for the two strains analysed (Fig. 8B). Only in supernatants of cultures grown in RAMGS of both strains, it was detected one prominent activity band with a Mw of approximately 170 kDa estimated with a calibration curve made using the protein standards.

Previously, there have been detected by zymogram analysis three Dsr designated DsxD (~ 308 kDa), DsrD (~ 180 kDa) and DsrI (~ 18 kDa) synthesized by L. mesenteroides subsp. mesenteroides ATCC 8293, upon growth in medium containing sucrose [49]. Thus, a blast analysis of the genomes of the F17 and F18 strains in comparison with that (CP000414.1) of ATCC 8293 was performed. The analysis revealed that, among of the three dsr genes present in the type strain, only exist in both Algerian strains the dsrD gene. Moreover, the F17 and the F18 genes shear among them 99.06% homology and they have 97.38% and 99.80% homology with the dsrD gene of the type strain, respectively.

These results support that the dsrD genes of the Algerian strains could encode the protein detected in the zymogram. Both genes have 4584 nt including the translational termination codon TAA. In silico translation of the genes indicate that both encode DsrD of 1,527 amino acids (aa) with a predicted Mw of 169.817 kDa and 169.802 kDa for the enzymes of F17 and F18, respectively. The Dsr are synthesized intracellularly and then they are processed and secreted. Thus, these enzymes have to be translocated through the cell membrane. In Gram-positive bacteria, the major route for exporting proteins across the cytoplasmic membrane is the general secretion pathway and proteins subjected to this type of export generally contain a N-terminal signal sequence, which is cleaved by signal peptidases upon translocation [50]. Therefore, as expected, the Signal P 6.0 program predicted a cleave site (VLG-DSS) for both DsrD between their 42 and 43 aa. Consequently, the mature extracellular proteins of F17 and F18 should have a Mw of 165.345 kDa and 165.330 kDa, in agreement with the estimated Mw (~ 170 kDa) of the bands detected by the zymogram analysis (Fig. 8B).

Furthermore, Fig. 8B shows that the active Dsr was synthesized at a very low level in RAMG medium. Thus, these results indicated that expression of the dsrD gene is induced, when sucrose is present in the growth medium. This behaviour is not a general characteristic, but it has been previously detected by zymogram analysis for other L. mesenteroides [51] and L. lactis [30] strains.

Finally, multi-align comparison of the dsrD genes from F17 and F18 with those of bacteria deposited in the GenBank database allowed the construction of a phylogenetic tree (Fig. 3 C and Supplementary Fig. S1C), which revealed that the dsrD gene of F17 has the highest homology with their homologs from L. mesenteroides DSM20486 (99.37%) and CBA350 (99.11%) strains. Also, F18 carries a dsrD gene with very high homology to that of F17 (98.99%) and to the others L. mesenteroides strains, being the highest homology with the gene of DRC1506 (98.98%) and the lowest with that of DSM20241 (91.69%). Furthermore, the homology of both F17 and F18 drastically decreased when their genes were compared with those of L. citreum, L. lactis and W. cibaria strains ranging the homology from 68.68 to 61.72%.

Analysis of metabolic fluxes from sucrose

DsrD, upon hydrolysis of sucrose, elongate the dextran polymer and can release (in addition to free fructose) glucose, a monosaccharide that can be used by L. mesenteroides as carbon source, and to generate among others compounds lactic acid [8]. Furthermore, L. mesenteroides can convert fructose into mannitol in a reaction catalysed by the mannitol dehydrogenase [52]. Provided that the active DsrD enzyme is extracellular, the metabolic fluxes of L. mesenteroides F17 and F18 in RAMGS were investigated by GC-MS analysis of the cell-free culture supernatants during growth (Figs. from 9 A to 9 F). Also, the evolution of the pH of the cultures (Fig. 9G and H) and the bacterial growth (CFU/mL) (Fig. 9A and B) were monitored. The F18 strain showed a better pattern of growth than F17 with a generation time of 76 min vs. 98 min and a growth rate of 0.79 ± 0.05 h⁻¹ vs. 0.61 ± 0.04 h⁻¹. In addition, the increase of biomass in 24 h incubation was higher for F18 than for F17, since although the initial inocula were similar (5 × 10⁶ CFU/mL and 3 × 10⁶ CFU/mL, respectively), at the end of the incubation the former strain reached a value of 1.4 × 10⁹ CFU/mL and the latter only 3 × 10⁸ CFU/mL.

Fig. 9.

Analysis of central carbon metabolism in F17 and F18 grown in RAMGS. The evolution of bacterial biomass, the pH of the culture and the presence of sugars and metabolites in the cultures supernatants are depicted. Symbols: ●, Viable cells (CFU/ml); ■, dextran; ■, glucose; ♦, sucrose; ▲, fructose; ■, mannitol; ♦, lactate and ●, pH. The values correspond to the average of three independent experiments. The results showed good performance of both strains, but with better behaviour of F18

Concerning to the metabolic analysis, again F18 showed a better performance than F17. The sucrose was added to the medium at a concentration of 30.77 ± 1.83 mM and after 24 h still 1.48 ± 0.88 mM of this disaccharide was detected in the supernatants of the F17 cultures (Fig. 9C), whereas only 19.09 ± 4.9 µM was observed in those of the F18 cultures (Fig. 9D).

Accordingly, the final levels of dextran production were 1.40 ± 0.24 mg/mL and 3.43 ± 0.3 g/L for F17 (Fig. 9A) and F18 (Fig. 9B), respectively. The two LAB generated transient extracellular accumulation of fructose during sucrose fermentation (Fig. 9E and F), then started to incorporate the fructose and the decrease of extracellular monosaccharide was accompanied by a release of mannitol, which accumulated in the supernatants of the F17 and F18 cultures up to 46.69 ± 0.30 (Fig. 9E) and 59.73 ± 7.24 mM (Fig. 9F), respectively, after 24 h of incubation. These results are consistent with those previously reported for L. mesenteroides, L. lactis, W. cibaria and Weissella confusa strains [16, 30].

Mannitol is a naturally polyol (sugar alcohol) that imparts a sweet taste without increasing blood sugar levels compared to sucrose and other simple sugars, because is poorly absorbed by the host intestine, thus, this polyol is recommended for the diabetic patients and it is an alternative sweetener for the chemical sugars used in industry [31]. Therefore, the capability of the two analysed L. mesenteroides strains to synthesise high concentration of mannitol, support their usage to generate healthy fermented products that could be consumed by diabetic patients.

In addition, both strains decreased the pH of the medium to a value of 4.6 (Fig. 9G and H), behaviour attributed to extracellular accumulation of lactic acid, and the highest lactate concentration was observed for F18 strain, which produced 76.57 ± 8.27 mM (Fig. 9H).

Surprisingly, the glucose present in the RAMG commercial medium (43.87 ± 2.58 mM) was not metabolised during the entire fermentation process and only slight variations in the levels were observed at the end of the incubation (47.79 ± 5.23 mM and 42.27 ± 0.88 mM in F18 (Fig. 9D) and F17 (Fig. 9C) samples).

Thus, the overall results showed that L. mesenteroides strains carrying DsrD, preferentially metabolise the added sucrose, instead of the glucose present in the medium, contributing to the increase of the biomass and to the decrease of the pH of the medium. Moreover, they revealed that the two bacteria produce high concentration of dextran, mannitol and lactic acid, good properties for LAB to be used to generate functional food. In this context, also the decrease of pH could provide to the end-products a specific aroma and to contribute to extend their shelf-life.

In vitro evaluation of the technological and probiotic capabilities

The bacterium L. mesenteroides has the Generally Recognized As Safe (GRAS) status of the USA Food and Drug Administration (FDA) and the Qualifed Presumption of Safety (QPS) status assigned by the European Food Safety Authority (EFSA), and it is widely used in food industry as starter or adjunct culture as well as probiotic.

Analysis of technological properties

The fundamental criteria in the fermentation bioprocesses is the ability of starter strains to acidify the matrix through carbohydrates fermentation, which generate organic acids giving to the end-product a specific aroma and extending their shelf-life [53]. Thus, the acidification potential of the two LAB as well as the evolution of the titratable acidity in milk due to them were evaluated in kinetic curves of pH measurement (Fig. 10A). The F18 strain showed a more rapid and strong acidification activity than the F17. Thus, in 14 h the former bacterium provoked a pH decrease from 6.42 to 5.98 pH units, and the titratable acidity values increased from 18.33 ± 0.58 °D to 30 ± 1.7 °D. Also, the appearance of curd and milk coagula took place after 36 h with F18 and 72 h with F17 characterized by the appearance of fissures. Consequently, the two strains showed a potential acidification profile in milk compared to other heterofermentative strains [54], and these results support the applicability of these strains in the elaboration of fermented milk products.

Fig. 10.

Acidification kinetics of F17 and F18 in skim milk medium supplemented with 0.3% of yeast extract and 2% glucose. The variation with time of pH and and dornic acidity (°D) are depicted. Symbols: ♦ or ♢ F17; ▲ or △ F18, respectively. The values correspond to the average of three independent experiments. Both LAB were able to acidify the tested medium, being F18 more efficient

Some milk proteins are considered as allergens for human beings. In this context, the LAB proteases target mainly caseins and whey proteins, and this types of proteolysis can effectively reduce their allergenicity and/or antigenicity [55]. Furthermore, the generated peptides are essential source for growth-stimulating aa. In addition, the catabolism of these polypeptides generate bioactive peptides during milk fermentation and contributes significantly to the development of some organoleptic properties in different fermented products [56]. Thus, the proteolytic activity of the F17 and F18 strains was tested in PCA-agar medium supplemented with reconstituted skim milk. A clear halo around the bacterial spots between 20 mm and 25 mm was generated by the two strains (Table 3; Figs. 11 A and 11B), indicating the presence of intra and extracellular proteases. These results are in agreement with those previously obtained for other L. mesenteroides strains [16, 34].

Table 3.

Technological properties of the L. mesenteroides F17 and F18 strains

Strain	Proteolysis (mm)		Lipolysis
	Skimmed milk		Olive oil			Tween 20
	1%	5%	1%	3%	5%	1%	3%	5%
F17	2.03 ± 0.1	2.5 ± 0.1	+	+	+	+	+	+
F18	2.0 ± 0.1	2.4 ± 0.1	+	+	+	+	+	+

Fig. 11.

Detection of proteolytic and lypolytic activities of F17 and F18. The bacteria were spotted on PCA-agar medium supplemented with 3% (A) or 5% (B) skim milk as well as on buffered MRSG-agar at pH 7.0, opacified with 0.5% CaCO₃ and containing 3% olive oil (natural lipid source) (C) or 3% tween 20 (artificial or lipidic source) (D). The clear halos detected around the inoculated spots revealed proteolytic (A and B) and lipolytic (C and D) activities of the two L. mesenteroides strains

In addition, the macroscopic visibility of milk coagulation hinges on the formation of a stable interwoven protein network. Acidic conditions usually promote it by inducing structural alterations in proteins, culminating in their aggregation and gelation. Conversely, substantial proteolysis results in enzymatic degradation of these extensive polypeptide chains into small peptides and free aa, rendering them deficient in their ability to construct the resilient, three-dimensional matrix required for a discernible coagulum. According to [57] the proteolytic potential in L. mesenteroides is attributed to the expressed protease encoded by a gene similar to those reported for L. mesenteroides ATCC 8293, J18, DRC0211 and BD1710 strains. The gene homology analysis using NCBI BLAST tools further confirmed that F17 has a putative protease coding gene (CDS: 502159–502890) with 100% identity with that (CDS: 505649–506380) of ATCC 8293 strain, which has also a homolog in the F18 genome (CDS: 500069–500803) with 99.59% homology.

In addition, it was analysed the lipolytic capability of the L. mesenteroides strains (Table 3). The detection of a clear zones surrounding the inoculated spots in buffered MRSG-agar medium supplemented with natural or artificial lipidic source indicated the lipolytic potential of the two strains (Figs. 11 C and 11D). These results are consistent with those obtained with others Algerian strains [16, 34]. Lipolysis is a complex metabolic process, which involves a vast array of enzymes encoded by numerous genes such as carboxylesterases, true lipases, and various types of phospholipases and, accordingly to the lipid molecule encountered, varying the enzymatic machinery [58]. Thus, the genomes of F17 and F18 strains were explored to detect genes encoding potential lipolytic enzymes. It was detected that the esterase/lipase coding gene of L. mesenteroides ATCC 8293 (CDS: 2014491–2015384) had a 99.66% homology with genes of F17 (CDS: 2022737–2023627) and F18 (CDS: 1958385–1959275). Also, the coding gene of the lysophospholipase L1 related esterase in ATCC 8293 (CDS: 670926–671816) has 99.66% homology with that (CDS: 667451–668338) of F17, and 98.65% with the homolog (CDS: 665332–666219) of F18. Finally, the homologous gene of ATCC 8293 (CDS: 738754–739665) has 91.75% homology with both CDS: 735281–736189) and CDS: 733154–734062 of F17 and F18, respectively.

In conclusion, the enzymatic activities described above complemented with the in silico analysis of the genomes make to predict that the usage of F17 and F18 for dairy fermentations could confer and enhance organoleptic properties to the end fermented product.

Analysis of probiotic properties

In vitro assays were performed to evaluate different properties of the potentially probiotic F17 and F18 strains (Fig. 12).

Fig. 12.

Response of F17 and F18 to various stresses. Survival of the bacteria after exposure to GIT stresses, hydrophobicity and heat shock as well as autoaggregation properties are depicted. For each individual treatment, the different letters indicate statistically significant differences between F17 and F18 with a one-way ANOVA analysis and a p > 0.05. The values are the average of three independent experiments. Means with the same letter are not significantly different. In most of the conditions both bacteria behaved similarly

First, different stress conditions were applied to mimic the GIT, since one of the baseline characters for probiotic bacteria is the capacity to survive in this environment. The obtained results revealed that the two strains exhibited non drastic differences in their survival rates under the different simulated GIT conditions. When the bacteria were exposed to the stress conditions of the stomach (pH 3.0 in the presence of pepsine at 3 mg/mL), it was observed a reduction of cell viability to 17.57% and 20.05% for F17 and F18, respectively. Also, the two bacteria demonstrated a progressive reduction in survival rate upon exposure to increasing concentrations of bile salt. Furthermore, F18 was slightly more resistant than F17, although with not a clear statistically significant differences, showing the former strain a minimum decline in viability to 70.87% and 38.35% upon exposure to 0.5% and 2.0% bile salt, respectively. Analysis of the bacterial response to the in vitro simulated stomach duodenum-passage (stimula) showed similar survival rate for both strains around 14.0% or 2.6% after 1–3 h exposure. This partial ability to tolerate the GIT stresses may be due to the release of some resistance protein such as the intra/extracellular osmotic regulation mechanisms of some molecules, since several studies revealed a bile salt resistance phenotype of Leuconostoc strains [59].

Also, as an indicator of the bacteria abilities to adhere to the intestinal epithelium, their cell surface hydrophobicity, microbial adhesion to hydrocarbon and autoaggregation ability were investigated. The two strains showed a high percent of hydrophobicity (31.72% or 30.15%, for F17 or F18). In addition, concerning to the autoaggregation ability, F18 exhibited a slightly higher values (53.23%) than F17 (44.71%). These results support a probiotic usage of the two Algerian LAB. However, it is noteworthy to mention that many factors can affect the probiotics-host interactions such as peristaltic movements, digestion and transit time, so it is difficult to estimate and predict the probiotic survival in the GIT tract environment only from in vitro tests [16].

Furthermore, the two strains showed a good resistance to phenol exposure being F18, the strain that displayed better behaviour. Thus, F18 showed a survival of 83.50% or 66.50% upon treatment with 0.2% or 0.4% phenol, whereas F17 displayed 65.80% or 51.78% survival, respectively. With regard, to resistance to moderate heat shock, not significant statistical differences between the two strains were observed, and F18, showed a 70.06% or 45.15% survival after 1–5 min exposure to 60 °C, whereas F17 displayed in the same conditions values of 64.30% or 34.79%. In addition, heating treatment to 80 °C showed to be a drastic heat shock for the two bacteria with only a 1.84% survival of F18 upon 1 min exposure.

Despite that Leuconostoc strains lack choloylglycine (bile salts) hydrolases (BSH) genes [60, 61], it has been suggested that several stress-related genes, beyond BSH genes can contribute to bile resistance. Following the methodology employed by Wu et al. [61], using PGAP annotation and blast tools from NCBI, we have identified several genes in both strains, which can be involved in response to the following stresses: acidic, alkaline, bile salt, oxidative, cold and heat shock as well as genes encoding proteins involved in aggregation, proteases, chaperones or universal stress protein (Supplementary Table S4). Thus, the presence of these genes in the genomes of F17 and F18 also potentially support the adaptability and tolerance of these strains to different stress conditions.

Concerning to potential antibacterial activity of the two L. mesenteroides strains, both cultures supernatants exhibited growth inhibition against to foodborne pathogen tested (data not shown). The cultures supernatants exhibited moderate to very strong growth inhibition of S. aureus ATCC 25,923, E coli ATCC 25,922, L ivanovii ATCC 19,119 and P. aeruginosa ATCC 27,853 foodborne pathogen in an indicator tested.

The antagonistic activity of LAB against foodborne pathogens is often due to the synthesis of secondary metabolites with antimicrobial potentials, such as organic acids, hydrogen peroxide, or proteinaceous molecules (bacteriocins).

Lack of growth inhibition of the pathogens was observed, when the culture supernatants of the L mesenteroides strains were tested after pH neutralization and treatment with proteolytic enzymes (results not shown). According to these results proteinaceous inhibitor agents were not detected and the inhibition zone in the first assay may be attributed to low pH or to the organic acids produced by F17 and F18 heterofermentative bacteria. Nevertheless, the genome analysis using BAGEL4 webserver [62], and protein-coding sequences homology determined with Blastp tool at NCBI and LABiocin database revealed that F18 possess genes with the capacity to encode peptide belonging to class IIb bacteriocin lactobin A/cerein 7B family (CDS 73839–74015).

These results indicate that further studies are required to detect the potential bacteriocin production by F18.

Test of safety considerations

As food adjuncts intended for human or animal consumption, various safety assessments of probiotics were investigated based to the current guidelines for safety assessment in food/feed to be considered suitable for the QPS status [63].

Thus, before future studies of their applicability, an evaluation of the safety properties of the two putative probiotic LAB was carried out through in vitro studies based on safety criteria including analysis of: (i) antibiotic resistance, (ii) potential BA production and (iii) potential hemolytic activity.

Concerning to antibiotic resistance, some LAB isolates have been shown to possess innate or acquired resistance to antimicrobials substances. As a result, it is vital to assess their antibiotic sensitivity, because any transmissible antibiotic resistance in a probiotic bacteria increases the risk of evolving acquired resistance in commensal pathogenic bacteria of the host body by horizontal gene transfer mechanisms. As starting point and as stated above, no acquired extrachromosomal antibiotic resistance was detected in the analysis of the plasmidic genomes using CARD and ResFinder webservers. Thus, to further analyse resistance of F17 and F18 to antibacterial compounds, it was investigated the susceptibility patterns of the two LAB to various antibiotics with different mechanisms of action, using the disc diffusion method assay and following the recommendations of the EFSA (Table 4).

Table 4.

Analysis of antibiotic sensibility of the L. mesenteroides F17 and F18 strains

Class of antibiotics	Mechanism of action	Antibiotic	Disc potency (µg)	L. mesenteroides F17		L. mesenteroides F18
				ZOI (mm)	Sensitivity pattern	ZOI (mm)	Sensitivity pattern
Aminoglycoside	Inhibition of protein synthesis	Amikacin	30	19	S	16	I
		Gentamicin	10	17	S	18	S
		Kanamycin	30	0	R	11	R
		Streptomycin	10	11	R	15	S
Phenicols		Chloramphenicol	30	26	S	32	S
Lincosamides		Clindamycin	2	33	S	30	S
Macrolides		Erythromycin	30	20	S	21	S
Others		Fusidic acid	10	19	S	16	I
Tetracyclines		Tetracycline	30	32	S	28	S
Penicillin	Inhibition of cell wall synthesis	Amoxicillin	30	33	S	21	S
		Amoxicillin/Clavulanic acid	20/10	29	S	22	S
		Ampicillin	10	30	S	23	S
1 st generation cephalosporin		Cefalexin	30	9	R	21	S
		Cefazolin	30	23	S	15	I
3rd generation cephalosporin		Cefotaxime	30	15	I	29	S
		Cefixime	5	0	R	0	R
Glycopeptides		Vancomycin	30	0	R	0	R
Polymyxin		Polymyxin B	300UI	11	I	12	S
Quinolones	Inhibition of nucleic acid synthesis	Ciprofloxacin	5	0	R	12	R
		Nalidixic acid	30	7	R	0	R
		Trimethoprim/Sulfamethoxazole	1.25/23.75	0	R	0	R

The mean zone of bacterial growth inhibition (ZOI) is expressed in mm. The pattern of inhibition is defined as: S Susceptible ≥ 20 mm, I Intermediate = 15–19 mm, R Resistant ≤ 14 mm

The analysis revealed that the two strains were susceptible to most of the tested antibiotics, according to the cut-off values of each antibiotic previously reported [36, 37]. However, they were resistant, in addition to vancomycin, to kanamycin, ciprofloxacin, trimethoprim/sulfamethoxazole, nalidixic acid, cefixime and streptomycin (only F17). These are no transmissible resistances, previously detected during the analysis of others L. mesenteroides strains performed by other researchers [64–66].

In addition, neither F17 or F18 showed any hemolytic activity, having a ɤ-hemolytic pattern, which is consider as safe phenotype compared to bacteria with α and β-hemolysis phenotype, which can have cytolytic effects on the host red blood cells.

Another safety issue is the BA synthesis, which is performed by many LAB [67]. The reason is, that when the BA accumulate in high quantities in the human body, some pathologic associated symptoms will occur, because they are precursors of molecules which cause inflammatory, allergen or immunological reactions [68]. Therefore, the development of fermented foods requires the use of a starter without or with precise and unambiguous BA production. Thus, F17 and F18 were investigated for the decarboxylation of L-tyrosine, L-ornithine, and L-histidine aa, which are the precursors of tyramine, putrescine, and histamine respectively, and any of the two bacteria was able to synthesise these BA (results not shown). In this context, it has to be stated that also predicted ORF with significant homology to BA coding genes were not identified by genomic analysis.

In summary, the above results support that the two LAB investigated in this work are safe L. mesenteroides strains, that can be used as starters for the development of fermented functional food.

Conclusions

In this study, two L. mesenteroides strains designated F17 and F18 were exhaustively screened following in vitro safety evaluations and probiotic profile analysis with a comprehensive investigation into their genomic features and fermentative behaviours. These bacteria displayed desirable technological properties notably their rapid milk acidification as well as proteolytic and lipolytic capacities. The ability of these bacteria to synthesise bioactive compounds such as dextran and mannitol during sucrose fermentation offers an alternative strategy for food biofortification using natural bioactive compounds. In addition, the two LAB produce riboflavin. However, the levels of the vitamin synthesised by them are lower than those reported for overproducing LAB mutants and consequently they do not seem to be good candidates for food biofortification with riboflavin. Both strains exhibited significant survival rates under in vitro simulated stomach duodenum and stress conditions with a high surface adhesion. In addition, none of the two strains showed any pathogenic properties, and they displayed a significant antimicrobial activity against foodborne pathogens. Analysis of the whole genome of the strains further confirmed their probiotic/safety profile and revealed the presence of genes linked to riboflavin and dextran synthesis as well as synthesis of protease and adaptive responses to stress conditions. In general, the properties and the behaviour of L. mesenteroides F18 indicate that this strain is more suitable than F17 for a future use as probiotic starter or adjunct for the development of fermented functional food. Future research will be focus on in vitro development of biofortified fermented dairy products involving this strain to explore its potential applications.

Abbreviations

aa

Amino acids

BA

Biogenic Amines

BHI

Brain Heart Infusion medium

CDS

Coding sequences

EFSA

European Food Safety Authority

FDA

USA Food and Drug Administration

FT-IR

Fourier transform infrared

GC-MS

Gas Chromatography-Mass Spectrometry

GRAS

Generally Recognized as Safe

GIT

Gastrointestinal Tract

QPS

Qualifed Presumption of Safety

LAB

Lactic Acid Bacteria

MRS

Man, Rogosa and Sharpe medium

MRSG

MRS containing 2% glucose

MRSG-agar

MRSG supplemented with 1.5% agar

MRSS

MRS containing 2% sucrose

MRSS-agar

MRSS supplemented with 1.5% agar

MSE

Mayeux, Sandine and Elliker medium

ORF

Open Reading Frames

RAMG

Riboflavin Assay Medium containing 2% glucose

RAMGS

RAMG supplemented with 2% sucrose

rt

Room temperature

TEM

Transmission Electron Microscopy

Authors’ contributions

Conceptualization, K.Z. and P.L.; methodology, A.F.Z, I.F. and M.L.M., software, A.F.Z. and I.F; validation, M.L.M; formal analysis, A.F.Z., K.Z. and P.L.; investigation, A.F.Z, I.F., and M.L.M.; writing—original draft preparation, A.F.Z.; writing—review and editing, K.Z., M.L.M. and P.L.; supervision, I.F., K.Z., M.L.M. and P.L.; project administration, P.L. and K.Z.; funding acquisition, P.L.

Funding

Open Access funding provided thanks to the CRUE-CSIC agreement with Springer Nature. This research was funded by the Spanish Ministry of Science, Innovation and Universities [grants RTI2018-097114-B-I00 and PID2022-136874OB-C31] and two internships of A.F.Z. supported by the CSIC [grant I-COOP COOPA20488].

Data availability

All data generated or analysed during this study are included in this published paper. The nucleotide sequence of the chromosome and plasmids of the investigated L. mesenteroides strains have been deposited in GenBank under accession number (from CP178852 to CP178856) for F17 strain and (from CP178911 to CP178913) for F18 strain.

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.